Revenue And Customs v Sunico A/S & Ors

This is a claim by Her Majesty's Revenue and Customs ("HMRC") for damages and interest arising out of an alleged unlawful means conspiracy to commit missing trader intra-Community ("MTIC") Value Added Tax (“VAT”) fraud involving exports and imports of mobile phones. The First Defendant, Sunico A/S ("Sunico"), a company incorporated in Denmark in the business of supplying mobile phones, is the vehicle at the centre of HMRC's claim.

The claim arises out of some 719 transaction chains constructed by HMRC occurring between August 2004 and January 2006, each of which HMRC alleges is an instance of MTIC fraud involving Sunico. HMRC claims that, as a result of the MTIC fraud represented by these transaction chains, it has suffered losses in excess of £40 million.

It was recognised at an early stage in this litigation that a trial of so many transaction chains would be unmanageable in view of the significant volume of documentation underlying each chain. At a case management conference on 25 May 2011 before Norris J, HMRC submitted that there should be a trial of liability based on a sample number of representative transaction chains followed by - if liability was established - a trial to determine the quantum of damages. This approach was opposed by the Defendants, unless HMRC agreed to provide disclosure in relation to all 719 Transaction Chains. Norris J ordered that there should be a trial of liability based on 26 chains which, the parties being unable to agree, he proceeded to select from the transaction chains upon which the overall claim is based. The Defendants' liability in respect of these 26 chains (which were called “the Sample Chains”) is the subject of this trial.

At a case management conference on 4 and 5 October 2012 before Warren J, summary judgment was granted in favour of the Ninth Defendant, Mr Nari Premchand, dismissing the claim against him in its entirety. Summary judgment was also given for the Eighth Defendant, Mr Dayal Dhalomal Shahdadpuri, dismissing HMRC's claim of conspiracy. HMRC's remaining claim against Mr Dayal Shahdadpuri pursuant to s. 423 Insolvency Act 1986 was subsequently discontinued with the leave of Warren J on 14 November 2012. Consequently, the Eighth and Ninth Defendants take no further part in these proceedings. HMRC's claim is thus effective against the First to Seventh Defendants only. I will collectively refer to them as “the Defendants", or, in some instances which are I hope obvious from the context, I will mean the First to Fifth Defendants only. Mr David Chivers QC, Mr Peter Shaw and Mr Tiran Nersessian represented HMRC before me, and Mr Abbas Lakha QC and Mr Graham Brodie represented the First to Fifth Defendants. The Sixth and Seventh Defendants did not appear and were not represented.

Sunico was incorporated in Denmark on 15 May 1996, its principal business being the trading of mobile phones. The defence asserts that Sunico has been an approved distributor for, among others, Samsung, Motorola, Siemens, and Panasonic. The Second Defendant, Mr Sunil Harwani (who without intending disrespect but purely for convenience sake I will refer to as “Sunil”), is a director of Sunico and it is accepted that he was a director at all times material to these proceedings and in particular, the period during which the transactions in dispute occurred. He also holds a 49% interest in the issued share capital of Sunico through a Danish holding company, Abascus Holdings ApS, the Fourth Defendant, which he owns outright. Mr Mangharam Harwani (who for the same reasons as with Sunil I will refer to as “Mangharam”), the Third Defendant, is Sunil’s father and holds the remaining 51% interest in Sunico through his Danish holding company, M&B Holdings A/S, the Fifth Defendant, which he owns outright.

The nature of Mangharam's role in relation to Sunico's business, and accordingly the extent of his liability in respect of HMRC's claim, is a subject of dispute between the parties. However it is common ground that Sunil is the controlling mind of Sunico when it comes to establishing Sunico’s liability or lack of it. In other words, if he is part of the fraud, then so is Sunico; if he is not, then Sunico is not either.

The Sixth Defendant, PT Naina Exim Indo (“PT Naina”), is a company incorporated in Indonesia. PT Naina entered into an agreement (the "Commission Agreement”) with Sunico, pursuant to which PT Naina was purportedly entitled (according to the pleadings and Sunil’s account given to the Danish tax authority (“SKAT”)) to a share of Sunico’s profits in return for referring customers and/or suppliers to Sunico. PT Naina has taken no part in these proceedings.

The Seventh Defendant, Mr Hashu Dhalomal Shahdadpuri, resides in Singapore and is the brother of the former Eighth Defendant. According to statements made by Sunil to SKAT, the Commission Agreement was negotiated with the Seventh Defendant. He too has taken no part in these proceedings. He has however participated in parallel proceedings commenced against him by HMRC in Singapore.

HMRC obtained a freezing order in this country against all the Defendants. A parallel order was obtained against the Seventh Defendant in Singapore. He unsuccessfully applied there to have that order set aside on the grounds that the claim amounted to an impermissible attempt by HMRC to bring a UK Revenue claim in Singapore. The ancillary freezing and attachment orders obtained against the Defendants in Denmark, Singapore and Hong Kong have given rise to much litigation in those jurisdictions. With the exception of PT Naina, challenges have been made by all the Defendants in those jurisdictions to the orders on the basis that they have claimed that HMRC’s proceedings are an attempt to enforce UK taxes and so, it is contended, offend against the private international law principle of non-enforcement of foreign Revenue debts. In Hong Kong, the application of the Seventh Defendant to strike out proceedings there was dismissed. Similarly the application of the Eighth Defendant in Singapore to strike out the claim against him in that jurisdiction was dismissed by the Singapore Court of Appeal. The Danish Defendants have made an application to the court in Copenhagen for an order that any English judgment that may be made ought not to be recognised in Denmark. The matter has been referred to the European Court.

HMRC's claim against all the Defendants is for damages for unlawful means conspiracy. Insofar as the claim concerns the Sixth and Seventh Defendants HMRC allege, in summary, that the terms of the Commission Agreement are not bona fide arms-length commercial arrangements and that the agreement was in truth a mechanism for the division of the proceeds of the MTIC fraud. On that basis, the Sixth and Seventh Defendants would be parties to the unlawful means conspiracy and personally liable as such for HMRC's alleged losses. It follows that the case against the Sixth and Seventh Defendants will only stand if the case against the First to Fifth Defendants succeeds.

This is not a case brought in the First-tier or Upper Tax Tribunal as to the propriety of denying a tax reclaim and the extent of the connection between the transactions and the alleged fraudulent tax loss. HMRC put their case squarely on Sunico being a party to fraud. This court therefore has to decide whether Sunico did sell or supply mobile phones into the Sample Chains in circumstances where there was a VAT loss occasioned by fraud to which Sunico was a party.

A result of the sample approach adopted in these proceedings is that there has been disagreement between the parties as to the scope of disclosure. Having narrowed HMRC's claim to the Sample Chains, Norris J made an order for standard disclosure on 27 May 2011. Norris J did not in terms specify in relation to which of the transaction chains disclosure ought to be given, although it seems clear that he envisaged that disclosure was only required in relation to the Sample Chains he had selected for trial. To order wider disclosure would be to defeat the purpose of limiting the trial to a sample in the first place.

At the case management conference before Warren J the Defendants objected that parts of HMRC's evidence strayed beyond the Sample Chains, incorporating references to other transaction chains in relation to which HMRC had not provided disclosure. Warren J observed that if HMRC sought to rely on other transaction chains as relevant to the trial of the Sample Chains (for example, to demonstrate dishonesty on the part of the Defendants) evidence of those transaction chains might well be relevant. And if it was relevant, HMRC would need to provide disclosure in relation it.

In the event, Warren J was unable to hear the Defendants' application and reserved the issue to me as the trial judge. The outcome of this application is set out in my judgment given on 27 November 2012, the details of which I need not repeat here, save that I note below the effect of this decision on the witness evidence served by both parties.

Witness statements were served from the following individuals:

On behalf of HMRC:

Mr Roderick Stone OBE, who also gave oral evidence. He has been employed by HMRC since 1974 and, as is plain from his witness statement, he has extensive experience of dealing with MTIC fraud, other forms of VAT fraud and associated money laundering. Between 2001 and 2004, he was the MTIC fraud regional co-ordinator for Southern England, where he was responsible for implementing the strategy to combat MTIC fraud, and since October 2004, he has collaborated and advised on the development and delivery of the MTIC fraud strategy at HMRC. Notably, he is the HMRC representative on MTIC fraud related project groups sponsored by the EU Commission Tax Directorate. At the outset of these proceedings, the First to Fifth Defendants applied to have Mr Stone's evidence excluded on two grounds. First, it was said that Mr Stone's evidence was essentially expert or quasi-expert evidence and, as no order for expert evidence had been made, it ought not to be admitted. Secondly, the First to Fifth Defendants objected to parts of Mr Stone's evidence in respect of which HMRC had not given proper disclosure. In summary, I held that although Mr Stone's evidence is what I would describe as multi-purpose common form evidence, in that it deals with MTIC fraud in general rather than specifically in relation to the pleaded issues in the case, I accepted that his experience means that it is sensible to entertain his observations about MTIC fraud insofar as they assist the Court. However, I ordered that certain paragraphs should be excluded on the basis that they contain Mr Stone's opinion of Sunico's trading, and other paragraphs should be excluded because they contain tables constructed on the basis of information which had not been disclosed to the Defendants.

Mrs Susan Ogburn, who also gave oral evidence. She has been an officer of HMRC since its formation in April 2005. Prior to that, she was an officer of Her Majesty's Commissioners of Customs and Excise ("HMCCE"), which she joined in 1974. She left HMCCE in 1983 but rejoined in 1991. She has headed HMRC's investigation into Sunico, a role which she inherited from a Ms Sarah Lang. It was apparent from Mrs Ogburn's evidence that the 719 transaction chains were constructed by a team at HMRC under the direction of Ms Lang, before Mrs Ogburn became involved in the investigation. In addition to Mrs Ogburn's witness statement, HMRC also sought to rely on an affidavit sworn by Mrs Ogburn dated 17 May 2010 in support of HMRC's application for a world-wide freezing injunction against Sunico. This purports to set out the entirety of HMRC's allegations against Sunico and thus addresses all 719 transaction chains. The First to Fifth Defendants sought to exclude this evidence insofar as it traverses beyond the Sample Chains in relation to which HMRC has not provided disclosure. In my judgment of 27 November 2012, I held that only those parts of Mrs Ogburn's first affidavit which are inextricably linked to an issue related to the Sample Chains and 12 additional example transactions upon which she relied were to be admitted. I should add that I am satisfied HMRC have provided proper disclosure in relation to these 12 examples.

Mr Peter Sawyer, who was not called for cross-examination. He is an officer of HMRC and part of the special investigation unit within HMRC which worked to construct the transaction chains upon which HMRC's claim is based. In particular, Mr Sawyer was involved in the investigations into Riff Trading Limited ("Riff"), Fresh 'n' Clean Limited ("Fresh n Clean") and ECS84.com Limited ("ECS84"), which are the alleged Defaulters in some of the Sample Chains. In his evidence Mr Sawyer incorporates and corroborates three affidavits from Mr Kirk Butcher, Ms Julie Ward and Mr Gordon Young, all of whom were formerly involved with special investigation work at HMRC in relation to Riff, Fresh n Clean and ECS84.com respectively.

Ms Heather Rowe, who was not called for cross-examination. She is an officer of HMRC and part of HMRC's special investigation into Fone Rack Cellular Accessories Limited ("Fone Rack"), another alleged Defaulter in some of the Sample Chains. Ms Rowe's statement exhibits and corroborates an affidavit from Mr Stephen Cooper, who also worked on HMRC's special investigation into Fone Rack.

Mr Michael Jarvis, who was not called for cross-examination. He also is an officer of HMRC, and was previously the investigating officer in respect of Panther Services (UK) Limited ("Panther"), another alleged Defaulter in some of the Sample Chains. He exhibits and confirms the contents of an affidavit he swore in relation to a winding-up petition and application for the appointment of a provisional liquidator in respect of Panther.

On behalf of the Defendants:

Mr Alex Marsden, who gave oral evidence. He is a Fellow of the Institute of Chartered Accountants in England and Wales and works in BDO LLP's forensic accounting department. Mr Marsden was instructed by the Defendants to provide a forensic accounting analysis of the Sample Chains as constructed by HMRC and, in particular, to assess whether any monies that Sunico received through the Sample Chains were the sums for which they invoiced, and/or whether Sunico received additional payments, either directly from customers or from third parties, in respect of any of the Sample Chains. He was assisted in the preparation of his witness statement by other staff at BDO LLP, who carried out their work under his supervision and control. His report draws on several key sources of information, including: a computerised copy of the Microsoft Navision CF accounting system operated by Sunico ("Navision") for the financial years ending 30 September 2005 and 30 September 2006 (the key years relating to the Sample Chains); atrial balance from Navision as at July 2004, August 2004 and September 2004; Sunico's customer accounts for 16 EU customers; selected bank statements and financial statements; and discussions with various Sunico employees. His evidence is in the odd category of being neither expert evidence nor direct evidence of fact.

HMRC did not formally object to the admission of Mr Marsden's evidence. Indeed, Mr Chivers accepted before Warren J that, to the extent Mr Marsden performed no more than what was essentially a number crunching exercise of the relevant data, his evidence would assist counsel in expediting the submissions on the evidence. It is, to borrow Mr Lakha’s description, a convenient mechanism by which to put the underlying material before the Court. I note however that both before Warren J and before me at trial, HMRC drew attention to the fact that parts of Mr Marsden's statement are in the form of expert evidence, in that they include his opinions and/or conclusions based on the data, making it comparable to the evidence of Mr Stone, to which the Defendants objected for similar reasons. There was a suggestion that, were the Defendants adamant that Mr Stone's evidence should be excluded in its entirety by virtue of being expert evidence, Mr Marsden's should also go out on something of a "tit for tat" basis. In the event, however, HMRC never made such an application and allowed Mr Marsden's evidence to stand. I therefore make the same point that I made in respect of Mr Stone's evidence that, to the extent Mr Marsden's evidence expresses his opinions on the documents, I discount it.

At the outset of proceedings, the Defendants had expressed their intention to rely upon evidence from Sunil and Mangharam (together “the Harwanis”). Witness statements from the Harwanis were served in August 2012 and, at the case management conference before Warren J, the Defendants were granted permission for the Harwanis to give oral evidence by video link from Denmark on the basis that HMRC refused to assure them that they would not be arrested if they came to the United Kingdom. On 12 November 2012, after HMRC had chased the matter of the video link, the Defendants' solicitors stated that both Sunil and Mangharam would in fact give their oral evidence live in person. Both the Harwanis were present in Court throughout these proceedings.

Upon the conclusion of HMRC's case, however, Mr Lakha announced that there would be no oral evidence from either of the Harwanis and as to oral evidence the Defendants would be relying only on that of Mr Marsden.

The reasons behind this dramatic change of position were the subject of considerable argument in closing. Mr Chivers invites the Court to draw adverse inferences against the Defendants from their decision not to call their two primary witnesses of fact. Mr Lakha submits that the decision not to call the Harwanis was taken in the light of significant weaknesses in HMRC's case which only emerged during HMRC's oral evidence and that, accordingly, no adverse inferences should be drawn. I address the force of these submissions, along with the authorities on adverse inferences, in detail below. I make only some brief observations here as to the practical effects of the withdrawal of the Harwanis' evidence on proceedings.

First, until the seventh day of the trial Mr Lakha had apparently intended to call the Harwanis. In opening he made extensive submissions on the evidence that he expected to adduce from them and, as invited, I read the statements of the Harwanis as part of my preparation for the case.

There was, accordingly, some debate as to whether Mr Lakha had "kept his powder dry" and whether it was possible for him to withdraw his clients' evidence at such a late stage. Tsavilris Russ (Worldwide Salvage & Towage) Ltd v R.L. Baron Shipping Co. S.A. (The "Green Opal") [2003] QB Adm. Ct. 523 was discussed. In the end, Mr Chivers did not pursue the point and the parties agreed to proceed provided that the Harwanis' statements and exhibits were removed from the trial bundles, Mr Lakha's opening skeleton submissions were updated to remove all references to this evidence, and I take no account of such references in the hearing transcripts or my own notes. This approach seemed to me to be a sensible way of progressing and I am grateful to the parties for their co-operation in this regard. Frequently a judge in this Court is called upon to put evidence out of his or her mind. As Tomlinson J noted in The Green Opal at 535:

Secondly, and on the same basis, the witness statements cannot be taken into account and my pre-reading of them must be disregarded: see CPR 32.5 (1).

Thirdly, as part of his investigation, Mr Marsden was provided with information about Sunico's business and financial performance by both the Harwanis; it was, therefore, hearsay evidence in the context of Mr Marsden's report. Following the Defendants' announcement that the Harwanis would not be giving evidence, Mr Chivers objected (both orally and in writing) that this meant HMRC would not have the opportunity to test that hearsay evidence in cross-examination of the Harwanis. This was something HMRC always envisaged that they would have the opportunity to do, which is why, he said, they did not object to the inclusion of hearsay in Mr Marsden's report when it was first served. It was understood that Mr Marsden’s evidence would be helpful because it was for the most part the subject of direct evidence which was to be given by a primary witness of fact, Sunil. The parties eventually agreed that those passages in Mr Marsden's statement which depended upon information provided by the Harwanis should be redacted so that the matter would be dealt with in that way.

Fourthly, Sunil's statement exhibited three signed witness statements from a Mr Simon Japes, served by HMRC and read in other proceedings before the First-tier Tribunal in relation to a company called Tricor plc (“the Tricor Litigation"). Mr Japes is a former employee of HMRC who worked as part of the MTIC fraud team. Two of his witness statements were served on HMRC by the Defendants in August 2012 and the third was served on HMRC more recently, but shortly after it came into the possession of the Defendants. It appears that these statements came to the Defendants somewhat by chance; they were not disclosed by HMRC. Following the withdrawal of Sunil's evidence, Mr Lakha made it clear that he nonetheless still intended to adduce Mr Japes’s evidence, which was to the effect that certain documents (“the Hawks documentation” discussed below) is inherently unreliable. Mr Lakha preserved his clients' right to deploy Mr Japes’s witness statements notwithstanding the fact that they were not, technically speaking, before the Court. This gave rise to a submission by Mr Chivers that Mr Japes's evidence should not be admitted, or at least only admitted to the extent that it was put to Mrs Ogburn in cross-examination.

I accept that HMRC is, like any litigant, entitled to require the Defendants to comply with the rules of evidence. In particular, I acknowledge the submission that Mr Japes's statements amounts to no more than hearsay and the Defendants have not served a hearsay notice in respect of them as is required under the Civil Evidence Act 1995.

However, I found HMRC’s standpoint on this issue an unattractive one. HMRC were adopting a position whereby they were seeking to exclude evidence in one court on technical grounds whilst simultaneously, and admittedly, seeking to rely upon that same evidence before another tribunal. Mr Japes’s evidence was proffered by HMRC in the Tricor Litigation in proceedings which are still on-going. Moreover, this same evidence (or at least the majority of it) has been in play in these present proceedings since August 2012. Mr Japes’s evidence is plainly what it purports to be; if its provenance was suspect, HMRC had plenty of time to object to it on reasoned grounds. However Mr Chivers eventually recognised that Mr Japes’s evidence was likely to assist the Court and he did not pursue an application for its exclusion. Nor did he find it necessary, as at first he said he might, to insist upon Mr Japes’s witness statements being formally proved by way of a witness statement from the Defendants' solicitor (who also had first hand knowledge of how they came into the Defendants’ possession), a concession which saved time and costs. Accordingly, Mr Japes’s evidence is before the Court, subject to questions of relevance and the weight to be attached to it as hearsay evidence.

Fifthly, much of the background to Sunico and its trading was contained in the evidence of the Harwanis contending at length that Sunico’s trading was a genuine business. This has two consequences. First, the withdrawal of the evidence means there is a limit to what I can take into account by way of context. Secondly, the issue of the legitimacy or otherwise of Sunico's business which had been raised in those witness statements necessarily became of much less relevance to the case.

Sixthly, there were detailed submissions as to whether any inferences should be drawn from the decision of the Harwanis not to give evidence.

Seventhly, I bear in mind that parts of the pleadings comprising the defences of the Defendants consist of positive averments and denials which are now unsupported by any evidence and are therefore liable to be struck out.

In the absence of the Defendants' main witnesses of fact, the witness evidence in this case was primarily adduced in order to put the voluminous documentary evidence underlying the Sample Chains before the Court in a manageable fashion. Accordingly, and in the absence of any expert evidence, much in this case turns upon my assessment of the documentary evidence in the light of the parties' respective analysis of it. As I have already noted, to the extent that the witnesses expressed their opinions on the documents they discuss, I have discounted their evidence. I also bear in mind that the views of Mrs Ogburn, although a senior employee of HMRC, do not bind HMRC.

One category of documents which has proved particularly divisive is the Hawk Documentation. This expression describes a body of documents contained in 10 freight files belonging to and obtained from a company known as Hawk Precision Logistics Limited ("Hawk"), one of the freight forwarding companies used by Sunico to ship consignments of mobile phones into the UK. These freight files contain release notes and transport documents relating to each consignment of mobile phones shipped by the freight forwarding company, and HMRC rely on these documents to show that Sunico was the supplier of consignments of mobile phones into some of the Sample Chains.

The files containing the Hawk Documentation were only disclosed to the Defendants, along with HMRC's intention to put these documents before the Court, some seven days before the trial. There was however already exhibited to the witness statement of Mr Stone a schedule (“the Grant Thornton Schedule") based upon the Hawk Documentation produced by Grant Thornton some of whose partners are the liquidators of Hawk.

At the beginning of the trial the Defendants applied to exclude the Hawk Documentation on grounds of unreliability. Mr Lakha advanced this application with the serious allegation that HMRC had misled the Defendants and the Court in relation to the Hawk Documentation and the Grant Thornton Schedule.In summary, Mr Lakha alleged that: (1) HMRC were aware that Hawk had been identified as a perpetrator and orchestrator of MTIC frauds as early as 2006; and (2) HMRC had been in possession of the Hawk Documentation since at least 2005, yet they told the Defendants and the Court that HMRC had only come into possession of the Hawk Documentation following a request to Grant Thornton to provide the documents.

The outcome of the Defendants' application is detailed in my judgment of 27 November 2012 on the issue. However, it is necessary that I set out the submissions on the Hawk Documentation again in this judgment, as my conclusions on those submissions have a bearing on the case generally. In addition, the way in which the Hawk Documentation has been disclosed to the Defendants and put before the Court causes me some concern.

In support of his first allegation -that HMRC was aware of Hawk's role in MTIC frauds- Mr Lakha referred me to the decision of the Court of Appeal in R v. Sandu (Monmohan Singh) [2006] EWCA Crim 606, in particular the summary of Hooper LJ at the beginning of his judgment at [4]:

Again, [14] shows that Mr Anderson, who was prosecuting counsel, had opened the case on the following basis:

The Crown was thus seeking to rely on the fact that there were no records of MSCoten, the company which played the role of the Defaulter (as explained below) in the MTIC fraud at Hawk. However, at a pre-trial hearing, Mr Anderson had successfully prevented the defence from knowing the full extent of HMRC’s investigations into the freight forwarders (including Hawk). Returning to [14]:

HMRC did not want to reveal that there were ongoing investigations into the freight forwarders so as not to jeopardise those investigations. However, Mr Anderson was forced to concede that the prosecution position had been misleading. New information had been provided to the prosecution team in the week before the trial showing that Ms Hooper and Ms Ayton of Hawk had been deceiving HMRC officers by providing them with supporting documentation for transactions said to have taken place on one day when in fact they related to another day.

As a result the prosecution in R v. Sandhu provided a revised disclosure statement in the following terms:

The Court of Appeal noted however that this revised statement was deficient. In particular, it failed to disclose the identity and (notably in the case of Hawk's managing director, Ms Hooper) seniority of relevant employees and it also failed to reveal the very close working links between the freight forwarders and the offshore EU companies. Further, in response to Mr Anderson's submission that there was no suggestion that Ms Hooper and Ms Ayton had been removing documents or inserting bogus documents in Hawk's files, Hooper LJ observed:

It was on this basis that the appeal in R v. Sandhu succeeded and, at [28], Hooper LJ sets out a table showing the admissions that ought to have been made by the prosecution at trial. For present purposes, the following admissions are the most pertinent:

Similar conclusions to those in R v. Sandhu can be found in the decision of Crane J in R v Uddin & Ors (2005) (unreported) concerning HMRC's "Operation Venison". In that case, the prosecution had sought to rely on witness statements from Ms Hooper and Ms Ayton exhibiting several documents seized from Hawk. Despite strong evidence against the defendants, a stay of proceedings was ordered because HMRC was found to have committed an abuse of process by not disclosing its suspicions of Hawk (and, additionally, the fact that it was protecting Hawk as a source of information) and, therefore, the unreliability of the documents it had placed before the court. In his judgment, Crane J was particularly critical of HMRC's investigation teams and the extent to which they failed to reveal the problems with the Hawk material. I note especially the following passage at [116]-[117]:

And, importantly, at [184]-[185]:

Finally, in closing Mr Lakha cited the VAT Tribunal decision of 2005 in Aircall Export Limited (Decision No 19185). A week before the trial in Aircall commenced, HM Customs and Excise served a CD on the Defendant containing some 5,200 pages of documents relating to Hawk. This CD was accompanied by a covering letter in which HM Customs and Excise stated that it did not consider that the existence of this documentation undermined its case, but recognised "that it might open up a line of cross-examination for [the Defendant's] counsel", (Aircall at paragraph 9). The letter also noted that there had been criminal proceedings concerning Hawk (the Tribunal were I assume referring to R v. Uddin) and that these proceedings had called into question the reliability of documents obtained from Hawk. Although HM Customs and Excise in Aircall had ostensibly disclosed the Hawk material for the benefit of the defendants, the Tribunal was nonetheless critical of the way in which the material had been produced:

Secondly, Mr Lakha alleged that HMRC had misled the Court in that the Hawk Documentation, disclosed by HMRC just days before this trial began, had in fact been in HMRC's possession and control since at least 2005, contained on HMRC's data storage system known as "Electronic Folder". It was suggested this was no oversight by HMRC. Mr Lakha submitted that HMRC's reason for not making disclosureat an earlier stage was that it did not want to acknowledge the inherent unreliability of the Hawk Documentation. Therefore, Mr Lakha said, HMRC asked Grant Thornton to provide the Hawk Documentation, so as to avoid having to admit to the Defendants and the Court that it already had these documents in its possession.

In support of this allegation, I was taken to correspondence going back to 2003 between HMRC, specifically Mr Stone, and Hawk. In this correspondence HMRC made it clear to Hawk that its documents were to be examined and copied pursuant to HMRC's enhanced verification programme. By around 2005 it was evident that there was to be regular reporting required of Hawk. I was also taken to sections of Mrs Ogburn's affidavit and evidence from Mr Young and Mr Cooper referring to specific Hawk documents of a similar type to the Hawk Documentation, without explaining their provenance. Finally, Mr Lakha pointed to an earlier form of Mr Stone's witness statement (dated sometime in June 2012 and sent to the Defendants by email, apparently by accident), which did not exhibit the Grant Thornton Schedule at all, although its existence was mentioned in the body of the statement. Mr Lakha asked me to infer from these matters that the 10 files obtained by HMRC from Grant Thornton were already held by HMRC.

In response to this allegation, Mr Chivers accepted that HMRC not only had the powers to go to Hawk and "uplift" documents, but that HMRC did in fact exercise these powers. He also accepted that Mrs Ogburn, and most likely Mr Young and Mr Cooper, exhibited documents which had been taken from Hawk in this way. However, he pointed out that this was the only evidence the Defendants raised in support of a very serious allegation and that, short of demonstrating that HMRC actually held the Hawk Documentation or showing that HMRC had uplifted the entirety of the documents held by Hawk, it was not possible to draw the inference posited by Mr Lakha that HMRC held the specific documents in this case. Moreover, Mr Chivers’s instructions were that only some 3,500 documents relating to Hawk were in fact scanned into HMRC's Electronic Folder for Hawk.

Mr Stone was cross-examined about the contents of HMRC's Electronic File on Hawk. He denied that HMRC had taken possession of a substantial part of the Hawk Documentation; he maintained that the substantial body of Hawk files were taken by Grant Thornton alone. Mr Stone did concede that as soon as HMRC became aware that Grant Thornton held the Hawk files in around 2008 or 2009 it would have been possible for HMRC to access that material, but this concession is irrelevant to Mr Lakha's allegation.

As I observed in my judgment of 27 November 2012, the Defendants have not provided sufficient evidence to support this second allegation. In order to draw the inference that Mr Lakha invited, I would need to find that HMRC and their solicitors consciously concealed the Hawk Documentation with the intention of springing it on the Defendants at a later stage, thereby putting them at a disadvantage. It was not open to me to draw such an inference without more specific evidence about HMRC's Electronic Folder, the enhanced verification procedures carried out on Hawk, and how the Hawk Documentation was treated. Absent bad faith, which I did not find, the question of disclosure has to be approached with proportionality and I was not prepared to go behind the disclosure statement provided by HMRC to the effect that it and its solicitors had carried out the appropriate, reasonable searches required under standard disclosure.

Mr Stone's evidence was that there was some degree of inconsistency within HMRC as to how uplifted documents were stored, particularly in the period prior to 2006. Notably, Mr Stone was somewhat vague when asked if there was any internal guidance as to what types of documentation should be uploaded to Electronic Folder. He said that before 2006 the question of what uplifted material should be uploaded onto Electronic Folder was very much left to the discretion of the individual officer. If material was not uploaded electronically, it would be stored in hard copy in his office. It goes without saying that, in a large organisation with no central database of all documents uplifted, in which several special investigation teams were working on different cases against different entities, documents may have gone missing or lain undiscovered. It is therefore possible (although this is a matter of surmise only) that the Hawk Documentation was stored elsewhere in the organisation and that the HMRC team preparing the case against Sunico were simply unaware of this. Short of observing that the procedures described by Mr Stone at least prior to 2006 may leave something to be desired for a public body engaged in the preparation of substantial litigation against trading companies, I draw no further conclusions from this issue.

I remain concerned by Mr Lakha's first allegation, however. In my judgment of 27 November 2012, I was prepared to admit the Hawk Documentation on the basis that, although there were questions of unreliability, the Hawk Documentation may not be unreliable for the purposes for which it was to be adduced by HMRC, namely to show that Sunico supplied into some of the Sample Chains. I was not prepared to rule out potentially probative evidence at that pre-trial stage.

In the light of the decisions in R v. Sandhu, R v. Uddin and Aircall, it is clear beyond doubt that HMRC was well aware, at least by 2006 if not before, that the evidence obtained from Hawk was inherently unreliable. In Aircall HMRC expressly acknowledged this unreliability (albeit late in the day) by providing material from Hawk to the defendant for the purposes of cross-examination. Yet, in the present proceedings HMRC did not disclose any of this information to the Defendants, instead apparently proposing to tender the Hawk Documentation and rely on it to show supplies by Sunico. At the very least, Mr Lakha suggested, the Hawk Documentation should have been disclosed with some sort of health warning in view of its treatment by the courts.

Mr Chivers did not however accept that were it not for the work by the Defendants' solicitors and counsel, the unreliability of the Hawk Documentation, discussed at length in the decisions of other courts, might well have gone unnoticed in this Court. His riposte was that the trial did not open until after the Defendants had made their challenge to the Hawk Documentation on grounds of reliability and that, accordingly, no Hawk material was put before the Court in circumstances where the Court was unaware of the potential involvement of Hawk in MTIC fraud.

However I remain troubled by the lack of any suggestion that HMRC would have raised the issue of the unreliability of its own accord, had the Defendants not made their challenge.

I accept the argument that HMRC, despite being a public body, enjoy the same rights as any litigant, in that they are not obliged to assist other parties in preparing their cases and are entitled to prepare their own case in the best possible way. Yet I have no doubt that HMRC were aware, as any litigant needs to be, that standard disclosure under CPR 31.6 includes material which adversely affect their own case. It is true that this does not necessarily include material which in itself does not adversely affect a case but may provide lines of inquiry leading to other information having a negative effect. However, the disclosable material in this instance directly casts doubt upon the credibility of HMRC's own evidence, which, indeed, HMRC appeared to recognise in Aircall.Therefore, having been reprimanded in the Crown Court, the Court of Appeal and the VAT Tribunal about its approach to documents taken from Hawk, it is surprising that HMRC did not at least caveat the Hawk Documentation when they disclosed it by reference to these decisions, leaving it to the Defendants to excavate.

Fortunately in this case, the credibility of the Hawk Documentation was flagged by the Defendants at the outset, so that this issue has been at the forefront of my mind throughout the trial.

The issues in the case are perhaps deceptively straightforward. As Mr Chivers put it, the case resembles Tristram Shandy in that for every hour of hearing there were at least two hours of preliminary argument.

Mr Lakha was disinclined to make any concessions about whether a fraud had been committed at all. As the case went on, however, he seemed to accept and indeed aver that there were underlying MTIC frauds. His case was that, if so, Sunico was caught up innocently without wrongdoing by Sunico or the other remaining Defendants.

The parties framed the case broadly according to the same two issues:

First, were the Defendants parties to an unlawful means conspiracy (the "Conspiracy Issue")?

Secondly, if the Court finds that the Defendants were parties to such a conspiracy, can HMRC prove that it has suffered loss in any of the Sample Chains? Mr Lakha added that a necessary part of this second issue must be to determine whether the Sample Chains have been accurately reconstructed by HMRC. This depends on the cogency of the evidence upon which the claim is based.

It follows from this formulation that the Sample Chains are only engaged when and if I find Sunico to be a party to the fraudulent conspiracy; if HMRC fails on the first question, its claim fails in its entirety and the Sample Chains become irrelevant.

The remainder of this judgment is structured as follows:

First, I provide a general overview of the common features of MTIC fraud as described in the authorities and in the evidence, and an overview of the elements and authorities on the tort of unlawful means conspiracy.

Secondly, I identify and discuss the parties' submissions on the Conspiracy Issue. In light of the decision by the Defendants not to call the Harwanis, this issue involves consideration of the arguments and authority on the drawing of adverse inferences.

Thirdly, I discuss the 26 Sample Chains by reference to the documentary evidence and the parties' submissions as to the accuracy and reliability of those chains, and I address the issues of causation and loss.

Finally, I consider the position of Mangharam and the Fourth, Fifth, Sixth and Seventh Defendants.

The problem of MTIC fraud has become endemic within the EU over the past 10 years, causing huge losses in revenue to member states. Mr Stone’s figures suggest that the total estimated losses to HMRC between 1999 and 2009 range between £14.93 billion and £27.93 billion.

At least until about 2007, MTIC fraud in the UK predominantly involved computer chips and mobile phones. The convenience of such goods to MTIC fraudsters is obvious; they can be easily shipped and are essentially fungible, in that they can be easily packaged and repackaged into different consignments, making them difficult to trace.As a result, the fraudulent trades can usually be carried out very quickly, with the fraudsters able to disappear before the tax authorities are able to respond.

When the VAT system is operating correctly, it should be tax neutral in its effect upon VAT registered traders, regardless of the volume of their trade. Any supply of goods or services subject to VAT made by a VAT-registered trader to another VAT-registered trader should be charged with VAT on the purchase price. The VAT element of the sale price must then be accounted for to HMRC as output tax by the selling trader. The output VAT paid by the purchaser may then subsequently be reclaimed as input tax, either when the purchaser itself receives VAT on a sale (which can then be off-set against the VAT it owes HMRC), or, if the purchaser does not receive VAT, or sufficient VAT, on its sales, by way of an input tax payment from HMRC. In MTIC fraud transaction chains, however, the output tax owing to HMRC is unpaid at some point in the chain, while further down the chain, a claim for input tax is made. The net effect is a deficit in the revenue collected.

The constituent elements of MTIC fraud have been set out in several authorities which the parties referred to in argument. Most usefully for present purposes is the outline provided by Blackburne J in Regalway v. Shillingford[2005] EWHC 261 (Ch) at [3]-[11]. In summary:

A VAT registered trader in one EU member state sells taxable goods to a VAT registered trader in another member state (such as the UK). The overall effect is that the importer does not suffer VAT on its import.

The UK importer, known in MTIC parlance as a “missing trader”, "hijacked trader" or “Defaulter”, sells the goods to another UK trader, known as a “Buffer”. When the Defaulter sells to the Buffer, it charges VAT on the supply. As the Defaulter only incurs output VAT on its supply, but has no input VAT on its own purchase from the EU supplier, it is liable to account to HMRC for the full VAT element on its sale.

The Defaulter then fails to file a return or account for the VAT to HMRC. Sometimes it will simply go missing, meaning HMRC is unable to pursue a claim for the VAT it is owed. Alternatively, the Defaulter will direct the Buffer to pay some or all of the purchase price to a third party or third parties (“Third Party Recipients”), thereby rendering the Defaulter unable to discharge its VAT liability. Such third party payments are central to the present case.

The first Buffer trader, or "First-line Buffer" will sell the goods on in the UK to another Buffer (a "Second-line Buffer") and may either receive payment from the Second-line Buffer and make payments up the chain of supply itself, or the First-line Buffer may direct the Second-line Buffer to make third-party payments up the chain.

There may be a number of sales to Buffers, in order to extend the paper trail and obfuscate the underlying fraud.

Eventually, a Buffer will sell to an exporter of the goods (the “Broker”). The Buffer may account to HMRC for VAT in the usual way, netting off the input tax on its purchase against the output tax on its sales. The amount of the VAT payable to HMRC by the final Buffer in the chain will represent the VAT on the Buffer’s profit or commission.

The Broker will then sell the goods to a purchaser in another EU member state. It will suffer no effective output tax on the export, but it will have an input tax credit in respect of its purchase from the final Buffer.

The Broker will then submit a (frequently very large) VAT refund claim to HMRC in respect of the input tax from its purchase from the final Buffer.

Quite often the same goods exported by the Broker will then be re-imported to the UK and the whole sequence will be repeated again. Hence the name “carousel fraud”.

It may be some time before the importer defaults on his obligation to account for the input tax he has charged and been paid. In order to avoid detection the transactions are usually effected, or said to have been effected, extremely quickly, often within the space of one day. More often than not, the transactions involve substantial sums of money passing down the chain.

Overall, if the fraud depends upon the use of third party payments to put the money beyond the reach of HMRC, the Third Party Recipients (usually overseas companies) will receive virtually the entirety of the purchase price for the goods, including the VAT element, although each party to the fraud creams off a small percentage of the VAT as it passes on the goods.

In R (on the application of Federation of Technological Industries) v. Customs and Excise Commissioners[2004] EWCA Civ. 1020, Jacob LJ summarised the distinction between MTIC carousel fraud and the other variety of MTIC fraud, known as "acquisition" fraud (at [17]-[20]):

Thus the amounts that are paid to Third Party Recipients pursuant to the sales made by a Defaulter will be greater than the amounts that have been invoiced to the Defaulter by the EU Supplier. Further, the value of the excess payments to the Third Party Recipients will normally be substantially equal to the VAT unpaid by the Defaulter. By way of example, I refer to an illustration provided by Mrs Ogburn:

An EU supplier may sell 1,000 phones to a Defaulter for £99 each (on which no VAT is chargeable), giving a total purchase price of £99,000.

The Defaulter will then sell the 1,000 phones to a Buffer at £100 each, charging VAT on the price (as the sale is within the UK). This gives a total purchase price of £120,000 (£100,000 for the goods, plus 20% VAT), meaning the Defaulter has made a profit of £1,000.

The Defaulter then directs the Buffer to pay it just £1,200, being the Defaulter's profit of £1000 plus 20% VAT.

At the same time, the Defaulter will direct the balance (i.e. £118,800) to be paid to a Third Party Recipient.

Thus, the Defaulter has procured an overpayment from the Buffer, which has been paid to the Third Party Recipient, substantially equivalent to the Defaulter's VAT liability.

The Third Party Recipient has therefore received a sum that lawfully ought to have been available to the Defaulter to pay to HMRC in discharge of its VAT liability.

In a lawful chain of supply, in which goods are imported from the EU and subsequently exported, the net VAT treatment would be substantially neutral in that: VAT is receivable by HMRC from the importer; VAT is repayable by HMRC to the exporter; and VAT is accounted to HMRC on the net profit made on sales between UK vendors and purchasers. By contrast, the net effect of MTIC fraud transactions is that: no VAT is received by HMRC from the importer/Defaulter; but HMRC nevertheless makes a repayment to the exporter/Broker.

HMRC's case is that Sunico was part of a wider conspiracy to defraud HMRC. It is alleged that it supplied the mobile phones which formed the subject of the Sample Chains and that it was a Third Party Recipient.

The essential ingredients of the tort of unlawful means conspiracy are conveniently summarised by Nourse LJ in Kuwait Oil Tanker Co v Al Bader [2000] 2 All ER (Comm) 271 at p.312:

The constituent elements of the tort are therefore: (i) the use of unlawful means; (ii) loss and damage suffered by the Claimant; (iii) pursuant to a combination or agreement between the defendant and others to injure which need not be the predominant purpose. I therefore now turn to these elements.

In Total Network SL v HMRC[2008] 2 WLR 711, the House of Lords held unanimously that the unlawful means employed by the conspiracy did not need not be independently actionable (aside from the conspiracy) by the Claimant against any of the conspirators. It was sufficient that the unlawful means involved the commission of a relevant criminal offence. In this respect,I note the words of Lord Walker at [94]-[95] following his review of the relevant authorities:

Words to similar effect can be found in the judgment of Lord Mance. After noting that caution must ordinarily be taken when considering the scope of the tort of conspiracy, he makes the following observation at [120]:

It is therefore clear, and the Defendants did not dispute, that the authorities establish that the VAT fraud perpetrated on HMRC by MTIC arrangements constitutes sufficient unlawful means upon which to ground an action for conspiracy. This is provided, however, that the claim is properly formulated in terms of loss.

Total Network concerned a claim for unlawful means conspiracy brought in relation to an MTIC fraud by HMRC against a Spanish recipient of payments involving 13 carousel frauds concerning mobile phones. The facts as alleged were sufficient to give rise to both the common law offence of cheating the Revenue and statutory offences pursuant to s.72 of the Value Added Tax Act 1994 (“VATA”). This raised the question of whether VATA provided a "total regime" in respect of the remedies available to HMRC, thereby eliminating its rights of action at common law. The House of Lords was divided as to the proper characterization of HMRC's claims. The majority of their Lordships held that there was nothing in the statutory scheme to preclude HMRC's pursuit of a common law claim against an overseas company upon whom the statutory remedies under VATA did not bite (Lord Hope and Lord Neuberger dissenting). At [134]-[135], Lord Mance delivered the opinion of the majority:

As Mr Chivers observed, Lord Mance's conclusion depends upon the claim for conspiracy being properly characterised as one for damages – i.e. for the losses caused to HMRC by the fraudulent conspiracy - and not simply as a claim for the VAT payments owed by the Defaulters.

As was the case with the Spanish defendant in Total Networks, Sunico is a company registered in Denmark and, accordingly, it has no liability under the UK VAT statutory regime. It is plainly a non-taxable person of the type which their Lordships in Total Networks were concerned should be subject to common law remedies, and HMRC's particulars of loss are properly formed in this regard. Of course, the question of whether HMRC has suffered damage in fact and, if so, whether such damage occurred as a result of the Sample Chains, is the second main issue in these proceedings, which I consider below.

Combination or agreement for a common purpose

The authorities establish that there must be an agreement between the conspirators, yet it need not be an express agreement. Indeed, an unlawful means conspiracy may be found to exist even where the conspirators have never even met or corresponded. I note in particular the familiar words of Mr Justice Fitzgerald in R v. Parnell (1881) 14 Cox CC 508 at 514 (approved by the Court of Criminal Appeal in R v. Meyrick and Ribuffi (1930) 21 Cr App R 94 at 99):

Nourse LJ developed the point further in Kuwait Oil, in which he said at pages 312-3:

What must be common amongst the conspirators is an intention to harm the claimant. Again turning to Kuwait Oil, Nourse LJ approved the following dictum of Oliver LJ in Bourgoin SA v Minister of Agriculture, Fisheries and Food [1986] QB 716 at 777 which suggests that where a party acts with knowledge of the consequences (i.e. that harm will be caused) it is not possible for that party to deny an intention to cause that harm:

HMRC's primary case is that Sunico was a party to the unlawful means conspiracy and, therefore, had knowledge of the conspiracy. Mr Chivers, however, put forward an alternative submission that blind-eye, or "Nelsonian" knowledge on the part of Sunico would be sufficient to render it liable as a conspirator (relying on Attorney General of Zambia v. Meer Care & Desai[2008] EWCA Civ 1007). In AG of Zambia at [21], the Court of Appeal confirmed that a party can be found dishonest either if it had actual knowledge of the relevant facts, or if it has "a clear suspicion" of those facts and "deliberately decided not to enquire in order to avoid having confirmation that [they] were so".

Furthermore, it appears from R v. Siracusa (1990) 90 Cr App R 340 that simply standing by and permitting an unlawful act to be carried out may be sufficient evidence of turning a "blind eye" and acquiescing to the fraud. In particular, I note that words of O’Connor LJ at 349 (cited with approval in Kuwait Oil at pages 311-312):

Mr Chivers also referred me to the helpful and comprehensive review of the authorities in this area by Briggs J in Bank of Tokyo-Mitsubishi UFJ, Ltd v Başkan Gida Sanayi Ve Pazarlama AS[2009] EWHC 1276 (Ch). As Briggs J observed in that case (at [842]), the extent of the knowledge may be of particular importance in situations where primary fraudsters invite others to assist them in their dishonest designs on a strict "need to know" basis. In the context of MTIC frauds, this may describe the relationship between Defaulters (or the parties controlling them), which are so to speak the "primary fraudsters", and the EU Suppliers and/or Brokers at the end of the transaction chains.

This question of the extent of knowledge required was challenged by the Defendants. Mr Lakha gave the example of an international phone supplier such as Nokia, which conducts business in the knowledge that MTIC fraud exists in the mobile phone market. He submitted that it would plainly be absurd if Nokia was found liable as an MTIC fraud conspirator purely on the basis that (i) it was aware of MTIC fraud in the market; and (ii) in the light of that awareness, it believed that some of the phones it sold might be used for the purposes of MTIC fraud. In view of this absurdity, Mr Lakha submitted that in order for liability to be established, there must be active participation in the unlawful means conspiracy, and not merely facilitation. He referred me to CBS Songs Limited v. Amstrad plc [1988] 1 A.C. 1013, a case concerning breach of copyright. The defendant company was sued on the basis that it had sold machines capable of making unlawful copies of recordings. In the House of Lords, the question was whether the party which facilitated a breach of copyright should be jointly and severally liable with the infringer of copyright. In dismissing the claim against the defendants, Lord Templeman observed the following at page 1058E:

Mr Lakha also referred to Credit Lyonnais v. Export Credit Guarantee Department [1998] 1 Lloyd's Rep 19, in which Hobhouse LJ approves the following dictum from Aldous J in PLG Research v. Ardon International [1993] FSR 197 at pages 238-239:

Notwithstanding that PLG Research was a case involving IP infringement, Hobhouse LJ emphasises that these principles are drawn from the general law of tort: they support the proposition that knowing assistance is not enough to found tortious liability. It follows, Mr Lakha submitted, that blind-eye knowledge would be insufficient to establish liability on the part of the Defendants.

There is some difficulty with Mr Lakha’s analogy with Nokia. Nokia would not be deemed to have known of a fraudulent transaction; it cannot be said that the only reasonable explanation of the transaction in which it is involved is that it is connected to fraudulent evasion. However, whatever the merits of the argument (and I make no finding about it), knowledge on its own was not the basis upon which Mr Chivers advanced the point. In addition to knowledge of the fraud the defendant must also have the requisite common intention to cause harm to the claimant. This is the missing ingredient from Mr Lakha’s Nokia example. Although a major international phone distributor may supply phones into the market in the knowledge that some of these phones may come to be used as instruments of MTIC fraud, in no sense can that distributor be said to share a common intention with MTIC fraudsters to cause harm to the victims of MTIC fraud. In short, the distributor is not privy to the agreement or combination to conspire.

Clearly, there can be no intent to harm where the defendant does not have knowledge of the conspiracy. At the other end of the scale, and following the dictum approved by Nourse LJ in Kuwait Oil at 777 (cited above), deliberate action taken in full knowledge of a fraud and its consequences can amount to irrebuttable evidence of intention. Somewhere in the middle then, there may be cases where blind-eye knowledge is sufficient where the defendant acts in consort with other conspirators but deliberately shields himself from the truth, so as to avoid discovering the intricacies of the fraud as in AG of Zambia . Mr Lakha suggested that "active participation" is what is required, but that does not cover situations where a defendant (e.g. a director of a company) stands by and does nothing to prevent the unlawful acts: see R v. Siracusa .

In the final analysis, it seems to me that one comes full circle and the necessary intention to cause harm and supporting knowledge of the fraud are discovered by examining the nature of the underlying agreement or combination between the alleged conspirators and the defendant's proximity to that agreement or combination. This approach is neatly summarised by Briggs J in Bank of Tokyo at [847], after his review of the authorities:

As noted above, HMRC's primary case is that Sunico was a conscious party to the unlawful means conspiracy. Furthermore, questions of blind-eye knowledge are rendered secondary in view of the fact that the Harwanis did not give evidence: had they denied knowledge of the alleged MTIC fraud on cross-examination, Mr Chivers may have found it necessary to deploy a case based on blind-eye knowledge with more force. In the absence of such oral evidence, however, there does not seem to me to be any evidence in the case from which questions of blind-eye knowledge come into play.

I now turn to the Conspiracy Issue. The approach to this issue shifted following the Defendants' decision not to adduce the evidence of the Harwanis. Before setting out the questions which need to be addressed on this issue, it is necessary to consider the parties' submissions and the authorities on the drawing of adverse inferences.

Mr Chivers began by referring me to the familiar four principles summarised by Brooke LJ in Wisniewski v. Central Manchester Health Authority ([1998] PIQR 324, at page 340:

These principles are especially applicable in cases which raise serious allegations of wrongdoing against a defendant. In such circumstances, there is authority to suggest that a failure to give evidence is likely to lead to an adverse inference. In Crawford v. Financial Institutions Services Limited [2005] UKPC 40, (citing the well-known decision of the House of Lords in Herrington v. British Railways Board[1972] AC 877) there were proceedings against the defendant for misapplication of the funds of a bank which formed part of a group of companies of which the defendant was the ultimate controller. The Privy Council said at [7]:

Commenting on such decisions not to give or call evidence, Lord Walker observed at [12]:

The effect of the drawing of an adverse inference from a party’s failure to adduce materially relevant evidence was usefully summarised by Brown LJ (as he was then) in Benham Limited v Kythira Investments Ltd[2003] EWCA Civ 1794 in the following terms at [30]:

Benham was a case concerning submissions of no case to answer and it seems to me the decision not to call the Harwanis was, for present purposes, very similar to a submission of no case. Mr Lakha's position was that HMRC's claim was so hopeless that the Defendants did not need to provide an explanation from the two witnesses who (at least in the case of Sunil) had an intimate knowledge of Sunico and its business. This was despite the Harwanis being present in Court throughout. As I observed, the Defendants were, in effect, putting all of their eggs in one basket. As Mr Lakha did not go so far as to make a submission of no case, it was not necessary to hold him to his election to call no evidence. To some extent he sought to have his cake and eat it by deploying Mr Marsden and the Japes evidence but declining to call his primary witnesses of fact.

In Benham at [28] Brown LJ emphasises that in order for an adverse inference to be capable of being drawn, the party seeking to draw the inference must have adduced some evidence which "establishes a case to answer". This of course ties in with the third principle in Wisniewski, which indicates that this evidence need only be "some evidence, however weak" (Wisniewski at 340).

These last two words – "however weak" – indicate that the evidence need only be minimal. Indeed Brown LJ in Benham at [28] adopting the dictum of May LJ in Hughes v. Liverpool City Council, emphasised that all that is required to give rise to an adverse inference is a "scintilla of evidence". Mr Lakha attempted to put a gloss on this conclusion, suggesting that Brown LJ was not in any way displacing the test set out in the third principle of Wisniewski, and that the Wisniewski test – i.e. that there is a prima facie case to answer – was a higher test. I am not persuaded, however, that there is a distinction to be drawn here. I accept Mr Chivers’s submission that the expression "scintilla of evidence" is of equal standing with the test of "some evidence, however weak" in Wisniewski.

What is true, however, is that the question of whether there is a case to answer does depend on the individual case and the allegations in question. If the Court is to draw adverse inferences, they cannot simply be of a general nature; they must be specific inferences in relation to specific pleaded issues. I am mindful that this is a case where very serious allegations of fraud have been made against the Defendants and, whilst this does not affect the standard of proof, it does have some bearing on my approach to the evidence and the burden on HMRC to prove its claim.

If I do accept HMRC has shown a prima facie case to answer, the effect of adverse inferences can be to strengthen that case. Furthermore, inferences may serve to swing the pendulum in the claimant's favour, as explained by Brown LJ in Benham at paragraph 30 (in the context of situations where the defendant elects to call no evidence):

A similar point can be found in Lord Lowry's judgment in TC Coombs v. IRC [1991] 2 AC 283 at [300], although his Lordship also notes that, where the defendant has a convincing explanation for his decision not to give evidence, adverse inferences will not arise:

The effect of these authorities is that I must approach the conspiracy issue in these stages:

First, has HMRC shown a prima facie case to answer that Sunico was party to the unlawful means conspiracy, supported by some evidence?

Secondly, if I find there is a prima facie case to answer, should I draw adverse inferences against the Defendants from the failure of the Harwanis to give evidence in answer to that prima facie case?

Thirdly, if I do draw adverse inferences against the Defendants, do those inferences tip the balance in HMRC's favour and demonstrate that Sunico was party to an unlawful means conspiracy on the balance of probabilities?

It is important to note Mr Lakha’s explanation for the Harwanis’ failure to give evidence. This was that in the light of Mrs Ogburn’s evidence what he termed the three “central pillars” of HMRC’s case had been undermined to such an extent that there was no need for them to give evidence. He said that the averment that the First to Fifth defendants were participants in an unlawful means conspiracy was based on three central pillars, namely:

The receipt by Sunico of sums in excess of their invoices.

Payment was made to Sunico in accordance with payment instructions.

The profit earned by Sunico was greater than that earned by the others in the transaction chains.

He submitted that in the light of concessions made by Mrs Ogburn and the evidence of Mr Marsden, Sunico received what it invoiced, no more and no less; there was no evidence that payments were made to Sunico in accordance with payment instructions, which were shams; the allegation made by Mrs Ogburn that the VAT paid by (or as directed by) the Missing Trader was paid to Sunico was accordingly untenable and Sunico’s profit thus resulted from the ordinary commercial profit earned by a trader who sells at a higher price than he buys. He submitted that the allegation that the suppliers were inserted into the transaction chains by Sunico is also untenable since HMRC’s evidence is that suppliers purchased mobile phones from a number of companies and not just Sunico alone. Further Mr Japes’s statement is that all UK companies involved in the frauds related to Sunico had accounts with First Curacao International bank NV (“FCIB”) with subaccounts in the names of the suppliers and this did not apply to Sunico. Accordingly, submitted Mr Lakha, the only matter relied on was the receipt of payments by Sunico from third parties rather than its customers, a matter which could not be the foundation for fraud.

Mr Lakha submitted that the perpetrators of the frauds were the second line buffers who held accounts in the names of continental suppliers with FCIB. He attempted to establish by reference to flow charts that funds were circulated by the third party payers in a manner bypassing Sunico completely, giving the false impression that Sunico was involved in the fraud. He attempted to show the commercial explanation for the fraud by reference to each of the 26 transactions; namely how the money obtained by the conspirators moved through FCIB and was divided up between them. I note however that Mr Chivers also showed, at least in relation to eight of the Sample Chains, how a mark-up was alleged to have been extracted by Sunico, being the difference between its sale price and the total revenue generated through the export sale price, corresponding to a proportion of the VAT lost by HMRC.

In answer to Mr Lakha’s main argument, Mr Chivers pointed out that HMRC’s pleaded case is not that Sunico received more than it invoiced but rather that the Defaulters caused to be paid to Sunico more than it was itself contractually obliged to pay to its EU supplier. The contention that Sunico has accounted for every penny does not answer this allegation or engage with the question whether Sunico’s trading was dishonest.

The second question, whether Sunico’s profit was other than an ordinary commercial margin was, submitted Mr Chivers, irrelevant to the issue. The particulars of claim make no reference to Sunico’s profit and there was no evidence as to the commerciality (in terms of pricing etc) of any of the transactions.

The crux of Mr Lakha’s submissions was, submitted Mr Chivers, the unreliability of payment instructions. However, that is an issue which has to be determined on the balance of probabilities. The Defendants have therefore chosen to take the risk that the Court will assume that because some payment instructions in other chains, not before the Court on this occasion, are unreliable, then all payment instructions in the chains which are before the Court must be assumed to be unreliable. The Harwanis have chosen not to give evidence on whether payment instructions were in fact followed in the present case, nor on their relationships with the companies specified in Mr Japes’s evidence.

The issue for the Court is whether Sunico supplied mobile phones into the relevant chains. In some cases, there is evidence that Sunico did supply goods into chains in respect of which it did not receive payment in the exact sum it invoiced. Thus, the fact of not receiving payment in the exact sum cannot be conclusive that the phones were not supplied into the chain. It follows that payment instructions were capable of identifying Sunico as the supplier, even if the instructions were not followed.

In my judgment there are numerous factors in the evidence which raise a serious question mark over the bona fides of Sunico’s trading. Each one considered by itself is not a badge of fraud. However, standing back and applying common sense I believe that taken together they do require explanation.

First, Sunico traded with parties which were companies of straw and which were participants in fraud. The sales followed a consistent pattern in each chain. Those companies sold the goods at a contrived loss, they traded for a very short period, they did not make any payments to Sunico and they both required their purchases to be delivered to the UK and purchased in sterling regardless of their location. Mr Lakha says that this is no different from any supplier of mobile phones and does not indicate fraud. However, no due diligence was conducted by Sunico in relation to the parties with whom it traded, notwithstanding the large sums involved, and notwithstanding the fact that it is evident that on some occasions Sunico advanced substantial credit to them. In other words, there were apparently no checks at all.

There is very little documentation in relation to Sunico’s customers; only the most basic company and VAT registration documents and a letter of introduction which contained no names of individuals, details of the type and quantity of goods which it was proposing to purchase and no evidence of the customer’s ability to pay for them. Notwithstanding these sparse details, Sunico immediately entered into substantial transactions very shortly afterwards.

A Part 18 Request served by the HMRC asked for the names of the individuals at any of the customer companies with whom Sunico dealt. No such information was provided, nor was it provided in any of the disclosed documents.

There is nothing in writing (notes, emails or letters) on any specific issue nor is there any general business correspondence. There are no requests for refunds, no queries, no requests for payment by Sunico (even where sums were obviously outstanding), nothing.

There are no statements of account, either maintained by Sunico or sent to the customer. Where it was a third party and not the customer who was making the payment one would expect the customer to receive a statement showing the state of the account. It is extraordinary in such circumstances that the accounts could be reconciled where there was a turnover of many millions of pounds, hundreds of transactions over a very short space of time and no payments between Sunico and its customer.

Payment was invariably made not by Sunico’s customer but by a third party with whom Sunico had no relationship. I agree with Mr Lakha that by itself that is not an adverse factor, but taken together with other matters such as lack of record-keeping, it gives rise at least to the suspicion that Sunico was laundering the VAT into phones. Sunico consistently received third party payments in respect of sales to particular customers (and a particular class of customer) in relation to a particular class of business and the failure to make inquiry into the reason for and the source of the payment requires an explanation. There was no written confirmation (and no note of any telephone conversation) as to which third party was to provide payment for the customer. The amounts paid by the third party are odd: sometimes they exceeded the amount due from the customer, sometimes they nearly matched, but did not quite correspond to, the amount due. In no case did Sunico seek to recover the amount of any shortfall or offer to repay any surplus. Mr Lakha’s explanation, that there was a running account, is insufficient in the absence of the lack of records or an explanation from the Harwanis. Payments were often made by the same third parties notwithstanding the change in identity of the customer.

The transaction chains appear contrived in that the phones passed between numerous parties usually within a single day. Some of the parties were obvious fraudsters and defaulters. The phones then went to brokers who exported them. The EU suppliers made a loss in every case; all the other parties a profit.

Sunico delivered goods at its instigation to the freight forwarders and was the person which determined when goods would be released down the chain. This is a classic role in MTIC fraud: see the observations of Blackburne J in Regalway v. Shillingford (above) at [30]-[40].

Sunico was the person which chose where the goods were to be kept. It allowed customers to “ship on hold”, allowing the phones to be physically exported prior to release.

Taking the figures from Mr Lakha’s skeleton argument, it seems that the original UK seller sold to a customer outside the UK at a discount of £4.50 per phone below Sunico’s price. Assuming that Sunico had purchased the goods at market price, it would then have to pay the costs of transport to Denmark (reflected in the price it paid) and back to the UK (reflected in the price it charged) so that its role does not make sense unless Sunico knew that the broker was to make a VAT reclaim.

There was no proper system of credit control as one would expect from a bona fide business. A pro forma invoice was issued prior to the rendering of the final invoice which was given a different number and typically issued after the phones had been released to the customer. The receipts were recorded on Sunico’s Navision software on the same day that they were paid into the bank account but the invoices were frequently only issued after the phones had been released so that Navision’s records were, while by and large accurate, by no means entirely accurate and sometimes understated the true amount of the customer’s indebtedness. The pro forma invoices and excel spreadsheets (which, unlike Navision, did not record payments against any specific dates) were different from the Navision record.

Mr Chivers took the Court through Chains 1 and 2 to demonstrate this. I am satisfied on the balance of probabilities that the excel spreadsheets must have been made up after the event. Sunico’s accountants posted journal entries after the year ended 30 September 2006. It is plain from Mr Marsden’s evidence that the adjusting entries appeared first in Navision, which he described as “the book of prime entry”. The spreadsheets (which bear the same entry numbers) must in my view have been made after that date and were inserted (a) so as to make the spreadsheets result in a nil balance and (b) so as to appear that they were made during the course of Sunico’s trading. It is also plain that handwritten figures on the pro forma invoices were not made contemporaneously.

Further, Sunico released goods to customers without any security for payment. For example, there were shipments to Hawk for a Portuguese company Pak Shine Import Export LDA (“Pak Shine”) on 18 November 2005 where the instruction “ship on hold” was deleted and “shipment on release basis” was inserted. That was the very first invoice to Pak Shine. Payment of that invoice was not made until 25 November so that credit was being afforded in the sum of some £693,000 to a wholly new customer.

Another example is in relation to Sunico’s customer Inram Limited (“Inram”). Sunico released its last Inram consignment on 30 August 2005, creating a debit balance of £1,623,920 on Inram’s account. Subsequently, the sum of £559,500 was written off. It seems uncommercial, and requires an explanation, that Sunico would grant this level of credit to an entity of no substance in circumstances in which payments were received from third parties about whom Sunico knew nothing and now apparently avers were part of a fraudulent chain. Further there is no evidence that Sunico took any steps to demand or recover the loss it apparently suffered.

I also note that fraudulent parties apparently paid Sunico sums which it invoiced for supplies. That in itself suggests that Sunico was a party to the fraud. As Blackburne J said in Regalway (at [7] in the context of an innocent Buffer),

It therefore seems to me that (a) there is a prima facie case for the Defendants to answer and (b) I ought to make adverse inferences against the Harwanis for their failure to give evidence to explain the matters I have mentioned.

The Sample Chains selected by Norris J are as follows:

the Transaction Chains involving Riff Trading Limited (“Riff”) numbered 11a, 12a, 17a, 34a, 34b, 59a, 74b and 96a;

the Transaction Chains involving ECS84.com Limited (“ECS84”) numbered 1a, 6a, 17a, 18a and 21a;

the Transaction Chains involving Panther numbered 9b, 10b and 20 (66a);

the Transaction Chains involving Fone Rack Cellular Trading Limited ("Fone Rack") numbered 2d, 14a, 17d, 29c, 45c and 100c; and

the Transaction Chains involving Fresh 'N' Clean (Wales) Limited ("Fresh N Clean") numbered 15, 29, 47, 63/64].

I must bear in mind that I have to decide whether the chains are accurately reconstructed on the balance of probabilities. No higher standard of proof is required. I take into account all relevant matters and not merely, as I might if the test were the criminal standard of proof, whether continuity has been proved conclusively so that I am sure beyond a reasonable doubt. I also bear in mind that where there is no conclusive evidence the Court’s task is to consider which of various probable explanations is more likely.

Mr Lakha objected to the integrity of the Sample Chains on two linked bases: first, he contended that they were not accurately reconstructed and secondly he contended that HMRC had not proved loss through denial of repayment of tax. He attacked HMRC’s evidence as unreliable in four ways to prove reconstruction of the chains, causation and loss. He submits as follows:

The deal logs maintained by fraudsters are unreliable.

Third Party payment instructions are unreliable.

The Hawk Documentation is unreliable.

HMRC could not prove loss either because in some cases HMRC could not prove that the alleged Broker had exported the goods or because the Broker’s claim for a tax reclaim covered both domestic and export sales and it was impossible to ascribe the claim to any particular transaction chain.

I deal with these objections in turn.

The method by which HMRC was defrauded depended upon showing a paper trail from import to export of the goods. As Lord Scott said in Total Network at 731 C-D,

It is thus in the interest of fraudsters to record paper transactions accurately. It does not lie in the mouth of a conspirator to say that the paper trail was created by known fraudsters, or was incomplete, or contained mistakes or anomalies.

In virtually every case Mr Lakha objects to the integrity of the Sample Chains at the Buffer level on the basis that HMRC are relying entirely or chiefly on deal logs (by which expression I also include day books or supplier lists) maintained by the Buffers themselves. I observe in passing that owing to the nature of MTIC fraud it is hard to see how a transaction chain could ever be reconstructed without recourse to Buffers’ records. There is no proper basis on which to conclude that deal logs are bound to be inaccurate or are inherently unreliable as a record of a sale, although the Court must of course be vigilant to check that deal logs are consistent. Whether corroboration is required, and how great the measure of such corroboration, will depend on the circumstances.

It is notable that there is no relevant discrepancy between deal logs of different companies in the various Sample Chains despite the fact that they were produced independently of each other. For example (and this is only an example as the same is true in most of the deal logs) in Chain 3 the deal logs of NSD UK Limited (trading as the Fone Box), Owl Limited and Chandi Diwar Limited match precisely.

Instances of missing invoices in sequences in deal logs (or day books or supplier lists) are in any event negligible in the context of the many hundreds of transactions recorded.

The fact that a deal log does not include a particular invoice number does not necessarily mean that a mistake was made. There are other plausible explanations for the omission, for example that the transaction did not result in a sale or was cancelled: see Chains 1 and 19. Even if there were mistakes that would not mean that the deal logs as a whole were inherently unreliable.

Further, the Court does not have to find matching invoice or purchase numbers. Its task is to decide whether the claimant has made out its case on the balance of probabilities. Thus the Court can properly rely on evidence of dates, quantities or prices showing likelihood that the deals matched the transactions. HMRC have done this at points in Chains 2, 4, 5, 9, 10, 11, 13 and 15 to 26.

In some cases (see Chains 19, 21 and 22) there is a discrepancy of 1,000 between the number of phones shown to have been purchased and sold by Buffers so that in each case this number of phones is unaccounted for. However in the case of all these chains there is further corroborative evidence. Thus in Chain 19 (a) there is probably no discrepancy as both deal logs contain 169 lines and the quantity of phones in each line is equal on the sale and purchase side, (b) even if there were a discrepancy, a plausible explanation is that a particular sale and purchase straddles the month end and (c) as there were no other sale or purchases of that phone in that quantity by the alleged Buffer Dhalomal Kishore trading as Movil 2000 on or around 5 August 2000 it is to be inferred that the transactions were for the same consignment. In Chain 21 the same type of response can be made to the objections. Also Movil 2000’s deal log shows 1,561 phones being purchased on 3 August 2005 and then being sold on in three batches on the same day to the same purchaser. The three batches total the 1,561 phones and the invoice numbers are sequential. There were no other sales or purchases of that phone in that quantity by Movil 2000 on or around 3 August 2005. It can therefore be inferred that the transactions were for the same consignment. In Chain 22 the fact that Movil 2000’s overall sales were 1,000 phones less than its purchases is not a ground for concluding that this particular consignment (purchased from Zain Communications Limited (“Zain”)) was not sold on to Dhalomal Ramchand Pte Limited. The deal log shows that the phones were sold on the day of purchase from Zain (5 August 2005) in three batches with sequential invoice numbers.

Mr Lakha submits that Sunico’s name was inserted in order to conceal the identity of the true fraudsters, so that Sunico was entirely unaware of the entries. I accept that it is impossible to prove a negative. However, in deciding on the balance of probabilities whether this was the case I take into account the question marks over the bona fides of Sunico’s trading to which I have already alluded and the fact that the Harwanis did not choose to submit to cross-examination and put their case in evidence that they were innocent victims of fraud.

Secondly, there is the issue of the third party payment instructions. The Defendants rely on Mr Japes’s evidence and their analysis of the fraud based on the movement of money from FCIB. However, if Sunico supplied phones into the chains, the issue arises whether it did so with the intention of furthering the fraud. Mr Japes’s evidence does not show that Sunico did not supply phones into the chain or that it was not paid for that supply. That is an open question which was not in issue in the Tricor proceedings. Sunico has given no disclosure of documents relating to the transaction chains considered in the Tricor proceedings.

Sunico’s own records show that it supplied at least two batches of phones for which it was not paid in full, evidenced by an invoice to Europecom SARL (“Europecom”) in which there was a part non-payment of £62,200 and the invoice to Inram in which there was non-payment of £559,500. Although there is no evidence of a third party payment instruction in these cases, it is likely that there was because in neither case did the payment received match the invoice.

In each case the third party payments were made by companies with whom Sunico had no commercial relationship. Although it is possible to identify some payments made by third party payers as attributable to specific transaction chains in the majority of cases the third party payer made multiple lump sum payments to Sunico which it is not possible to link to particular transactions: see Chains 3, 7, 9 to 13, 17, 19, 20 and 23-26.

Sunico’s case is that although payment from third parties was in round sum amounts it rendered invoices to its customers. However statements of account disclosed by Sunico (such as from Inram, Interaxis LDA, Chartres Polka Spolka ZOO (“Chartres Polska”) and Eurosia Communications SO (“Eurosia”)) do not identify how the lump sum receipts are applied in payment of particular invoices. It is also wholly unclear how lump sum receipts relate to the handwritten annotations on Sunico’s pro forma invoices. In its defence Sunico maintains that it received written notifications from its customers of third party payments but no such notifications have been provided either on disclosure or by SKAT. As I have already said, this is in itself a suspicious circumstance.

The Defendants rely heavily on Sunico documents which appear to demonstrate that no consignment of the model of phone which it is said to have supplied in Chains 7, 11 and 13 was sold in the relevant period. However, in Chain 7, reliance is placed upon Sunil’s witness statement, now withdrawn. In any event, in all three cases it is reasonable to infer that payment instructions were followed because of substantial payments made to Sunico (by NSD, and Simplestar Limited respectively) and the pattern of trading and format of the payment are comparable to Chain 1, where there is evidence by way of invoices that Sunico supplied the phones into the chain.

I give full weight to the undoubted facts that Hawk was implicated in fraud and that the Hawk Documentation ought to be regarded with a jaundiced eye. However it is incontrovertible that Sunico sent hundreds of millions of pounds worth of phones through Hawk and was paid for them by fraudsters. Hawk undoubtedly received and despatched goods. That is plain from the important fact that the material from Hawk frequentlydovetails with documents produced by Sunico’s freight forwarders Kuhne & Nagel in Copenhagen whose integrity is not in issue. A paper trail was created to extract money from HMRC. Hawk used the actual receipt of goods from Sunico to sit at the head of the paper trail. There is a direct link between Sunico and the material used by the acknowledged conspirators. Much of the material may be inauthentic in that it does not reflect genuine economic activity; the issue on causation is whether Sunico, in the 26 chains, did indeed supply the goods at the head of the paper trail.

It is notable that release notes and other Hawk documentation is frequently corroborated by other information: see Chains 4, 5, 17 and 23. I was taken through Chain 2 in considerable detail to demonstrate the interaction of documentation and I observe that in that case the Hawk documentation corroborates and dovetails entirely consistently with documents from Kuhne & Nagel and from Jyske Bank in Copenhagen, Sunico’s Bankers.

I also accept that much of the Hawk Documentation is peripheral for any issue other than that of causation, that is to say, working through the deal chains. It is not relevant at all, save possibly in relation to Chain 14, on the issue of whether Sunico was a party to a fraud.

Again I remind myself that the standard of proof is the balance of probabilities. I should start by saying that I reject Mr Lakha’s argument that HMRC’s case “piles probability on probability” with the effect that the standard of proof in relation to the Sample Chains becomes a very high one indeed. The court has to decide a number of issues of fact, deciding each issue on the balance of probabilities. What the court does not do is say that if there are three issues to be decided and each is decided as to a 51% probability, the probability that they have all been surmounted is very small indeed, half of a half of a half.

I accept Mr Chivers’s analogy with The Creutzfeldt-Jacob Disease Litigation- Groups A and C Plaintiffs [1998] 54 BMLR 100 in saying that Mr Lakha’s approach is the wrong one and that the proof of causation of loss is not an ‘all or nothing’ one. In that case the issue arose whether certain plaintiffs had received a lethal injection of CJD before or after 1 July 1977 as, if it was after that date the defendant would not have been liable in negligence. The plaintiffs had each received numerous injections any one of which would have been lethal if it had contained CJD. Morland J said (relying on excerpts from the speech of Lord Bridge in Wilsher v. Essex Health Authority [1988] 1 AC 1074 at 1087) at 102-3,

Some of the VAT repayment claims show that in the relevant period the trader making the claim had carried out both domestic and export sales. Mr Lakha thus argued that where there is no direct evidence of export by the Broker in one of the Sample Chains HMRC is unable to demonstrate that it made a relevant VAT repayment in respect of the chain in question.

Of course, the facts in Creutzfeldt-Jacob were very far from the present case. However I agree with Mr Chivers that in those chains in which the quantum of the exports by the alleged Broker was greater than the quantum of domestic sales, and was large enough to cover the claim, it may be assumed that on the balance of probabilities that the relevant chain fell into the export rather than the domestic category. This applies to Chains 3, 7, 8, 15, and 17.

In many cases (see Chains 3, 7, 8 and 14 to 18) the Defendants’ objection was that there was no direct evidence of export of the goods by the alleged Broker. However, it is reasonable to infer that the goods were sold by the Broker; the sale of phones was after all its business. It is further reasonable to infer that they were exported because the Broker’s VAT return for the relevant period showed a greater amount of sales elsewhere in the European Union than domestic sales.

Chain 16 is an exception. The VAT return of S & I Electronics Plc shows that domestic sales in the relevant quarter (to 11 05) were greater than export sales in the ratio of approximately 5.3 out of a total of 10. In their written closing submissions HMRC’s counsel say as follows:

S & I seems to be a Broker and it also seems to be the final link in Chain 16. If that is correct, it must have been intended to make a tax claim from HMRC. The amount of the tax claim is easily enough to cover the tax which the Defaulter, Panther Services, did not pay. It was also the Broker in relation to Chain 17. However, in the light of the firm statement above and the fact that export by the Broker is a matter of inference only, HMRC is on the balance of probabilities unable to demonstrate that it made a relevant VAT repayment in respect of Chain 16.

HMRC have conceded that they cannot claim in respect of two of the chains, namely Chain 10 and Chain 12. With respect to Chain 10 HMRC accept that there is insufficient evidence of one link, namely a sale by Worldtech Solutions Limited to Sprint Communication Systems Limited because it is not possible to establish model, consignment size, unit price, invoice number or any other identifying feature of the sale which is evidenced only by a reference in Worldtech’s schedule of sales. As to Chain 12, it is conceded that the VAT reclaim of input tax by the Broker, FoneShops Limited, was denied so that no loss was suffered.

Both Mangharam and Sunil say (now without any supporting oral evidence on which they could be cross-examined) that Mangharam was not involved in the phone trading side of Sunico’s business. However that side of the business was huge and it defies belief that Mangharam had no knowledge of it. There was some dispute between the parties as to whether Mangharam was a director of Sunico; the dispute hinged on matters of translation and Danish law which I cannot resolve. However, it is evident from Sunico’s accounts that Mangharam held a position of authority at Sunico, even if he was not the Managing Director as alleged. The notes to Sunico’s accounts describe Mangharam as well as Sunil as having “deciding influence” in relation to related party transactions. Mangharam’s name appears on company transactional documents (for example authorisation to the Bank to pay Sunico’s supplier in relation to mobile phone transactions) and his own website publicity portrays him as active in Sunico’s affairs.

It therefore seems to me that there is evidence of Mangharam’s involvement in Sunico’s conspiracy. I add to that the fact that Mangharam did not give evidence either as to facts which he could have explained, or as to his allegedly minor role in the business, and on a Wisniewski basis the balance falls well and truly in favour of HMRC in relation to the 23 Sample Chains.

The Fourth and Fifth Defendants respectively hold 49% and 51% of Sunico’s share capital. All three companies are under the direction and control of Sunil and Mangharam. Sunil is the sole director of the Fourth Defendant; and it appears that Sunil, Mangharam and a Mrs Beena Harwani are directors or hold a similar position in relation to the Fifth Defendant. M&B’s accounts record Mangharam as having a deciding influence on the Fifth Defendant’s affairs by reason of his share ownership.

The knowledge of Sunil and Mangharam of Sunico’s participation in the conspiracy is attributed to both the Fourth and Fifth Defendants. Those companies through their shareholders agreed that Sunico would be used as the corporate vehicle for fraudulent trade and through which dividends would be paid.

In the accounting year ended 30 September 2005 Sunico made profits after tax of some D Krone 48.5m (some £5.5m). Of that profit Sunico paid a dividend of DKr 40,000,000 (some £4.5m) in total to the Fourth and Fifth Defendants. Thus a substantial part of Sunico’s profits was paid by way of dividend to these companies.

In my judgment the Fourth and Fifth Defendants are jointly and severally liable with the other liable Defendants for HMRC’s loss in relation to the 23 chains.

The Commission Agreement bears the date 20 December 1999 but it could not have been executed at that time (although it seems that there was an initial verbal agreement at that time) because Sunico did not have its current name at that time. It is believed that the Commission Agreement was executed in about 2002. It was negotiated on behalf of PT Naina by the Seventh Defendant who was its primary contact.

In his interviews with SKAT Sunil characterised the relationship between Sunico and PT Naina as a partnership which had “ led to the largest increase in Sunico’s turnover ”.

Sunil also said that there was continual telephonic contact between PT Naina and Sunico and that PT Naina would receive running telephone updates as to the progress of Sunico’s business.

However the arrangements between the two companies had numerous uncommercial and contrived elements. Thus,

There is nothing recorded in writing, no faxes, emails or notes of telephone conversations. Save for letters enclosing commission statements there are very few letters between Sunico and PT Naina or the Seventh Defendant.

It is alleged that there was a final indebtedness of $9.7m owed by Sunico to PT Naina in respect of commission. However there is not a single letter of demand, let alone any threat of proceedings. There is merely one letter listing the invoices and showing the outstanding amount.

Sunil told SKAT that all referrals were keyed into a computer database but no copy of that database or any printouts were provided on disclosure by Sunico or anyone else.

The letters enclosing the commission statements are addressed “Dear Sir” although there was supposed to have been continual telephone conversations and the parties knew each other well. There was a chain of existing relationships between them.

It appears that Sunico, Sunil and Mangharam may have attempted to commit a tax fraud on SKAT in that it was said to SKAT that all payments to the former Eighth Defendant made under the Commission Agreement represented legitimate payments to PT Naina of business expenses. However SKAT was not told that approximately £2m of this payment was paid back to Mrs Beena Harwani, Mangharam’s wife. This must on any basis have remained Sunico’s money but Sunico did not mention it in its affidavit of means sworn for the purposes of the freezing injunction. The balance of the commission payments was transferred to a company called Tallister Services Inc (“Tallister”) and to the Seventh Defendant. Large sums of money were sent through Sunico to Hong Kong and divided up into three parts: one for Sunico, one for the Seventh Defendant and the third, largest part, for an unknown person, seemingly Tallister.

The amount of the commission in the commission statements sent to PT Naina in the period corresponding to the Sample Chains is virtually all accounted for by sales to the 16 European Suppliers to the Defaulters in those chains, including Inram, Interaxis, Eurosia, Europecom and Chartres Polska. In circumstances in which each of the EU Suppliers seems to be little more than dummy entities it is hard to see how, without any other explanation, commission arrangements for introductions to them can be other than an arrangement to share the proceeds of a fraudulent conspiracy.

The only real evidence from Sunico on the subject is that which Sunil gave to SKAT. On the basis of this, it is evident that there was some arrangement whereby Sunico would benefit from contacts supplied by PT Naina. There was no purpose in the Commission Agreement at all unless Sunico was indeed receiving something to its advantage. HMRC assert that the introduction was to parties necessary to conduct the fraud and that this corresponds with the explosion in Sunico’s turnover. Again, there is no explanation to the contrary.

It is notable that neither PT Naina nor the Seventh Defendant has taken any part in these proceedings at all. They have neither served a defence nor given disclosure. I am not satisfied by their absence and I do find that there is a case to answer against them Accordingly I am entitled to, and do, draw adverse inferences under the Wisniewski principle. Accordingly I find that they too are jointly and severally liable to HMRC.

I therefore find that HMRC has made out its case in relation to 23 of the 26 Sample Chains; that is to say all 26 Chains with the exception of Chains 10, 12 and 16.

I am however conscious of the fact that the Sample Chains are only engaged as a result of my finding that Sunico was a party to the fraudulent conspiracy. There is a broader context to this litigation in that HMRC have some 693 alleged transaction chains waiting in the wings and a decision how to proceed with those chains may well depend on the findings I make. It therefore seems to me that I can and should state my finding that all the remaining seven defendants were parties to the conspiracy even though loss has not been shown in respect of three of the Sample Chains.