Mercury Tax Group Ltd & Anor, R (on the application of) v HM Commissioners of Revenue & Customs & Ors

The Information put before Judge Spencer consisted of a formal part in five short numbered paragraphs (I-V), deposing to Mr Preston’s suspicion that a serious tax fraud had been committed and stating why he believed the searches to be necessary; together with a further forty pages setting out the grounds for his suspicion. It is important to note that no copy documents of any kind were exhibited. The grounds were in five sections. I need not be concerned with sections 3 to 5, which deal with the premises in respect of which the warrants were sought and the details of their terms. Section 1 is entitled “Introduction”. Paras. 1.1 to 2 are purely introductory. Para. 1.3 gives a very broad general summary of the Scheme, in terms which I need not set out. The section then proceeds:

“1.4

Although the Mercury scheme was theoretically capable of achieving the effect it sought (and in that sense is not fraudulent in design) I suspect that it has been fraudulently implemented. I further suspect that this fraudulent implementation was in part envisaged from the outset by Mercury. The successful operation of the scheme relied upon the execution of a series of transactions effected by means of formal legal documents. I suspect that, for all users, it was envisaged from the outset that at least one of the steps would not be executed properly, but that documents would be presented to HMRC that purported to show that it was so executed. I believe that this was to have been done deliberately and that the acts were to have constituted an attempted fraud on HMRC. I suspect that this fraud was envisaged at the outset for all scheme users.

1.5

For at least some of the scheme users, I suspect that certain of the documents (including the documents at 1.4 above) have been fraudulently altered. I suspect that these alterations have involved the removal of signature pages from draft documents that had already been signed and their substitution into documents purportedly effecting similar, but entirely different transactions. The purpose of these frauds had been to demonstrate to HMRC that transactions that were not properly executed have been so executed. Again, I suspect this had been done deliberately and that the acts constitute an attempted fraud on HMRC.

1.6

Additionally, at least one document has been submitted to HMRC recording events that did not take place. I suspect that this has been done deliberately and in an attempt to increase the changes of the scheme being accepted by HMRC.

1.7

The scheme users were to have been the prime beneficiaries of the scheme via the claimed reduction in their tax liability. However, the professional firms involved in the design and implementation of the scheme also benefited through fees and interest payments from the scheme users and I suspect individuals within these firms to have been responsible in the first place for conceiving of a scheme that was to be implemented fraudulently as described at 1.4. I also suspect that they were responsible for carrying out the falsifying the documents as described at 1.5 and for submitting the document at 1.6.

1.8

I do not suspect the individual scheme users of having perpetrated these frauds themselves, and, it is conceivable that they had no knowledge of the frauds described at 1.4 and 1.6. For the frauds falling within 1.5 above, however, I believe that the effect of the fraudulent alterations was so profound the individuals affected must have been aware of this before their tax returns were submitted. I suspect that these individuals knew that documents had been fraudulently altered. I suspect, however, that they nonetheless submitted their returns in this knowledge and in the hope of gaining a reduction in their tax liability to which they knew they were not entitled.

1.9

The aggregate tax losses claimed by the scheme users totals £16,069,148, and the tax loss as a result of the suspected frauds is estimated to be more than £6m.”

Read by themselves, those paragraphs are pretty opaque; but in the light of the following section it can be worked out that para. 1.4 refers to the erroneous sequence provided for in the chronology; that para. 1.5 refers to the signature pages; and that para. 1.6 refers to the minutes of the meeting of 4 December 2003.

Section 2 is entitled “Details of the Suspected Fraud”. Subsections 2.1 to 3 set out the machinery of the Scheme in unexceptionable terms. Subsection 2.4 is entitled “Suspect Fraud”. It runs to 27 paragraphs, over eleven pages. It can be summarised as follows.

Paras. 2.4.1-9 set out the grounds for Mr Preston’s suspicion that the signature pages on the three key documents as signed by Mr Grisay were transferred from the earlier draft to the final documents. The case is put in summary at para. 2.4.1 as follows:

“The first evidence that leads me to suspect fraud in the implementation of the scheme derives from an examination of the documents submitted to the AAG [i.e. HMRC’s Anti Avoidance Group] in support of the claim made by Alain Grisay. I suspect that, in the case of Alain Grisay at least, [the Trust Deed, the Option Agreement and the Sale and Purchase Agreement] were fraudulently created. They are not therefore legally binding, but the documents have been presented to HMRC to give the contrary impression. I suspect that this was done in a deliberate attempt to deceive.”

Paras. 2.4.2-4 explain the change from the use of the 2003 to the 2004 stock and the changes in the amounts initially settled into the Trust. Para. 2.4.5 explains in elaborate detail the differences between the draft and final versions of the three key documents: such detail was of course necessary given the decision not to exhibit them. In relation to each Mr Preston records the opinion of the Government Chemist that the signatures on the draft and on the final version are identical. In relation to each he concludes:

“I suspect that Mr Grisay did not sign [the document]. I suspect that the signature page from the earlier draft deed was inserted into the later version that had been updated via a word processor.”

Mr Preston then says, at para. 2.4.6:

“The suspect fraud then involves the insertion of the signature page from an earlier, draft document. It is made possible by the practice of signing draft documents in advance of transactions.”

He goes on to give a further reason for believing that this was so. Mr Preston’s “suspicion” that all this was done because of the time constraints under which the scheme had to be carried out are further elaborated at paras. 2.4.7 to 9. Para. 2.4.9 concludes:

“I suspect that the planners anticipated that there would be insufficient time for all scheme users to have properly signed and executed the option deed in the required manner and, instead, sought and obtained signatures in advance of the transactions being carried out.”

As it turns out, of course, none of this is in fact controversial. The Claimants freely acknowledge that Mercury proceeded in the way “suspected” by Mr Preston. Mr Mitchell submitted to me that the effect, and indeed the intention, of the elaborate manner in which Mr Preston sets out the grounds for his suspicion is to create a quite spurious air of mystery about what was in truth an obvious fact. I agree that these paragraphs do have a somewhat Sherlock Holmes flavour; but it must be recalled that Mr Preston did not know what Mercury might or might not accept and was anxious to “show his workings” in a way which would satisfy the Court’s scrutiny. His repetitious use of the phrase “I suspect” reflects – I suspect – no more than a wish to stick as closely as possible to the language of the statute.

Para. 2.4.10 of the Information reads as follows:

“The chronology in Mr Clayton’s documents confirms these suspicions. This document was e-mailed to Mr Clayton by Michael Strutt of Pantheon [a firm which had introduced Mr Clayton to Mercury] on 10 November 2003 and I suspect it was prepared by Mercury. It sets out the steps in the scheme in chronological order and states explicitly that the option agreement is to be “executed” by the client (step 12) before the gilt strips were purchased (step 15). For the reasons set out at 2.4.9 above it would not have been possible for the deed to have been genuinely “executed” before the strip had been required, and I believe that the document envisages the mere signing of a draft document. I suspect that this document is an accurate record of the real sequence of events, and that the planners must always have envisaged asking for documents to be signed in advance of the transactions taking place.”

Thus the significance placed by HMRC on the chronology was not so much that the erroneous sequence of events which it embodied was itself deliberately fraudulent but that it was further evidence of the practice of obtaining signatures to draft documents, which is elsewhere described as “the principal fraud”.

In paras. 2.4.11-25 Mr Preston sets out a number of other matters on which he relies as supporting the suspicion that Mercury had adopted a general practice of obtaining signatures to draft documents and then transposing them to the final version. These included the results of an examination of the key documents by a forensic scientist: Mr Mitchell’s comments referred to at para. 22 above apply equally here, but again I do not believe that HMRC were seeking to create any unfair prejudice by seeking to prove so laboriously what they say happened. In view of the fact that it is common ground that Mercury did transpose the signature pages, I need not try to summarise the various points made. However, one point of particular importance concerns whether clients were told of the change from the 2003 to the 2004 stock. Mr Preston referred to the information given by Mr Kowalishin. At paras. 2.4.22-23 he says this:

“2.4.22

Once the decision had been taken to proceed with the June 2004 strips the scheme users ought to have been made aware of this and to have signed and executed revised documents. Mr Kowalishin says that he did become aware at some point that the scheme had not proceeded with the December 2003 strips, but he does not now recall precisely who made him aware of this or when this happened. The e-mails retrieved by Mr Kowalishin are highly significant here in what they do not include. Although the e-mail search was carried out for the whole of the period 1 November 2003 to 31 January 2005 inclusive the last e-mail that has been obtained from the search is dated 21 November 2003. If Mr Kowalishin had been made aware that a different financial instrument was being used and that revised documents would be required then I would have expected at least some references to this to have appeared in e-mail traffic, whether between Mr Kowalishin and Mercury or Mr Kowalishin and the other scheme users. No such e-mails have been recovered, however.

2.4.23

I suspect that the reason for this is that, as for Mr Grisay, those responsible for implementing the scheme decided not to seek properly-amended documentation. I suspect that, instead, they decided to falsify documents in order to make it appear that they had been properly signed and executed.”

It is clear from this passage that the Court was being invited to infer that clients had not been told about the change in the stock to be used; and that this was indicative of an underhand approach.

At paras. 2.4.26-27 Mr Preston deals with the question of the minute of the meeting of 4 December 2003. At para. 2.4.26 he makes the points which I have summarised at para. 15 above. He then continues, at para. 2.4.27:

“I suspect that the minute records events that did not take place. I further suspect that it has been included deliberately in the material submitted for Messrs. Grisay, Clayton and Meadon because Mercury believed that this would increase the chances of the scheme being accepted by the AAG. The Schroder & Co papers demonstrate that it had already been approached by Mercury by 20 November 2003 and contained no suggestion that any other potential purchaser was ever envisaged. I suspect, however, that the scheme planners may have been anxious to give the impression that other purchasers had been considered as it was widely believed in the tax profession at the time that (following WT Ramsay Ltd v CIR – 54 TC 101) introducing some element of “uncertainty” (over, for example the identity of a purchaser) into an avoidance scheme would reduce the chances of a successful challenge by HMRC.”

This section is headed “suspects”. It sets out the various individuals whom Mr Preston suspects of involvement in the frauds. I need not summarise this in any detail. Para. 2.5.1 reads as follows:

“I suspect the professional firms involved in designing and implementing the scheme to be responsible in the first instance for all the frauds described at 2.4 above. I suspect them to have encouraged all the scheme users to sign deeds purporting to execute transactions when they could not be validly executed, and to have created and submitted to HMRC the document at 2.4.26 in the hope of improving the chances of the claims to tax relief being accepted. I also suspect that the same firms were responsible for carrying out the substitution of pages from earlier draft documents into later ones in order to give the impression that the later documents had been properly signed and executed. None of these frauds can have been carried out unwittingly. Neil Masters, the Managing Director of Mercury, is a solicitor and I suspect that he was fully aware that all the acts were fraudulent.”

While he goes on to acknowledge that KB were fully involved in the implementation of the scheme, he goes on at para. 2.5.3 to say that he suspects that “responsibility lies primarily with individuals within Mercury”. It is at para. 2.5.6 that he refers to the substitution of the signature pages as “the principal suspected fraud”.

I have a transcript of the proceedings before Judge Spencer in Leeds on 8 November 2007. These show that the Judge had been supplied with the papers at the beginning of the day. He told Mr Preston, who was representing HMRC, that he had not had time to read them and offered to hear him either at 4.30 pm or on the following day. Mr Preston plumped for 4.30 pm. When the hearing resumed at that time the Judge said that he had read the documentation, and - if I may say so - it is clear from his following observations that he had made a careful attempt to identify the gist of the alleged fraud. He asked Mr Preston to tell him, so far as possible, “in two or three sentences … what the essence of the fraud is”. Mr Preston answered:

“We believe that the scheme was conceived using one particular set of gilt strips, and we believe that the scheme users signed documents with that first gilt strip. We think that later in the day, for reasons that we can speculate about, they decided not to use the first gilt strip, but to use a different gilt strip.”

The exchange continued:

“Q. The first gilt strips are the ones that mature in December 03? - A. That’s right.

Q. And then they use the ones that mature in June 2004? - A. June 04. We believe they were running close to the wire effectively, if they hadn’t got the scheme through before the maturity date then essentially the scheme wouldn’t have worked, so later in the day…

Q. Why not? - A. Well the essence of the scheme, the losses created by the granting of the option. The option only has a value if it can be exercised before the strip pays out, and so essentially they needed to allow themselves sufficient time for that to be done. So I think…

Q. So they missed the bus? - A. I think so.

Q. Yes, is your understanding? - A. It is, and we think that once they decided to go with the June 04 strips, for reasons we suspect are probably to do with, again one speculates, embarrassment, at any rate, we believe rather than have the scheme users sign a fresh set of documentation we believe that they removed the signature pages from the documents that the scheme users had already signed.

Q. Yes. Well, which of the potentials do you see as being responsible for, as it were, ringing the 04 documentation for the 03 documentation? - A. Well, in theory, it could have been individuals within Kleinwort Benson, or it could have been individuals within Mercury Tax Strategies. On the basis of the financial incentive, I think we think it is more likely that it was individuals within Mercury. … ”

There was then some discussion of other matters, after which the judge said:

“Q. Well, look, I have looked at the legislation in as far as you have provided me with 20C(1), and I have to be satisfied on information on oath that there is reasonable ground for suspecting that an offence involving serious fraud has been committed, and evidence of it is to be found on premises specified in the information. At the moment, on what you tell me, it seems that the parties having set up this arrangement, that these are the persons offering the tax avoidance strategy, having for whatever reason got themselves unable to use December 03 maturing securities, substituted securities maturing in June 04, right, have I correctly understood? - A. Yes, yes.

Q. And that was not the deal which the scheme users thought they were getting, we understand? - A. Mm that’s right.

Q. And your position is, so far as the scheme users are concerned, when they are presented with a tax return to sign, and it says “security maturing June 04”, they should have thought “blow me over, I signed something that said Dec 03”? - A. Exactly, your Honour.”

Having reached that point, the Judge indicated that he was willing to issue warrants in relation to the business premises and Mr. Masters’ home but not in relation to the private addresses of the clients. As to that, he said that he was concerned, apart from other matters, by article 8 considerations. There was then further discussion of the nature of the frauds alleged and in particular of paras. 1.4 to 1.6 of the Information in the course of which Mr Preston said this:

“Sorry, your Honour, it may be I have not expressed things as clearly as I ought here. The 1.4, the matter described at 1.4 is the perhaps narrow issue that I think there was a step in the scheme that it was always envisaged would be implemented incorrectly, that’s to say the scheme design document envisaged that the option agreement would be signed before the gilt had actually been bought. I believe that Neil Masters, who’s a solicitor, ought to have known that that was wrong. I don’t necessarily think that the scheme users knew that that was wrong. The issue at 1.5 is the substantive fraud, the one that we discussed earlier, the issue of the substitution, and that’s the one I believe the scheme users ought to have been aware of.”

In connection with the question of what “scheme users” - i.e. Mercury’s clients - knew or did not know Mr Preston repeated in some detail the points which he had made at paras. 2.4.22-23 of the Information about what Mr Kowalishin had, or had not, been told about the decision to change the strips. In the end, the Judge was persuaded to issue the warrants in relation to the private addresses of the clients as well.

Judge Spencer did not deliver any kind of judgment or summary statement of his reasons. It would have been better if he had done so – see the observations of McCowan LJ R v Southwark Crown Court, ex p. Sorsky Defries [1996] Crim L R 195, quoted by Kennedy LJ in R (Energy Financing Ltd.) v. Bow Street Magistrates Court [2006] 1 WLR 1316 ([2005] EWHC 1626 (Admin)), at para. 10 (pp. 1320-1). But some aspects of his thinking can be inferred from his exchanges with Mr Preston; and the absence of reasons does not in any event invalidate the decision.

I also have a transcript of the hearing before Judge Hillen, but this contains nothing further of any materiality.

The Claimants’ case, is as I have said, (a) that the material put before the Court (and, therefore, also before the Board at the approval stage) did not justify a reasonable suspicion that any serious tax fraud had taken place and (b) that in any event HMRC did not make full and frank disclosure.

It is convenient to consider separately the questions whether there was in fact reasonable ground to suspect the flaws in the implementation of the Scheme which HMRC have raised; and secondly the question whether, if so, there was reasonable to ground to suspect that those flaws were the result of dishonesty.

WAS THERE REASONABLE GROUND TO SUSPECT THAT THE IMPLEMENTATION OF THE SCHEME WAS FLAWED ?

I will consider the three alleged flaws in turn. I take first the issue of the substituted signature pages, since, as Mr Preston made clear both in the Information and in his oral submissions before Judge Spencer, that was the principal fraud suspected.

Mr. Mitchell submitted that there was nothing wrong in the procedure adopted. He submitted that the governing principle in law was that a contract in writing could be effectively altered after signature by a party provided that the person making the alteration had the authority of the party in question to do so (or, if he did not, if his act was subsequently ratified). He said that that was established by the decision of the Court of Appeal in Koenigsblatt v. Sweet [1923] 2 Ch 314. That, he submitted, was what had happened in the present case. So far as the final version simply filled in the blanks in the earlier version, those changes were plainly implicitly authorised: a party signing a document containing such blanks must envisage that they will be completed, and he will be bound so long as the words inserted fell within the scope of what he could reasonably have expected. – he referred to United Dominions Trust Ltd v Western [1976] QB 513.

Mr Mitchell accepted that that particular reasoning was not available as regards the substantive changes which I have set out at para. 12 above, which do not correspond to any blanks or square brackets in the signed draft. But in those cases, he submitted, there was material from which HMRC could and should have concluded that clients accepted the changes in question and implicitly, if not expressly, authorised any alterations to the final version to give effect to them. As regards the most significant of those changes - that is the change from the 2003 to the 2004 stock - Mr Mitchell submitted that documents already in HMRC’s possession showed that clients were told in terms, prior to the execution dates of the Option Agreement and the Sale and Purchase Agreement, that 2004 stock would be used. On 25 November 2003 KBPB sent Mr Grisay a formal facility letter in relation to the loan of the £950,000 which he was asked to, and did, sign and return by way of acceptance. This clearly stated, in the first paragraph, that the purpose of the loan was “to assist you with the purchase of £1,000,000 5% Treasury Stock June 2004 Gilt Strip.” What is more, Mr Grisay wrote to KBPB on the same day (though no doubt the wording was supplied to him by them) asking them to

“… take this letter as my authority for you to draw down the facility and utilise the proceeds thereof, together with sufficient funds available in my account maintained with you, to purchase £1,000,000 worth of 5% Treasury Stock June 2004 Gilt Strip. ”

It can reasonably be inferred that these letters were in standard form and that similar letters had been sent to and by the other clients. Although KBPB’s letter does not in terms point out there had been a change in the identity of the gilt strips or ask for authority to change the descriptions in the Option Agreement and the Sale and Purchase Agreement, the change was nevertheless evident and Mr. Mitchell submitted that the client must be taken to have authorised any alterations in the final form of the documents to give effect to it. Mr Mitchell had no equivalent submission about the other changes which I identify at para. 12; but in relation to the amount settled under the Trust Deed it could perhaps be said that the client must have appreciated that he was writing a cheque for £3,000 not, £2,510.

As a fallback, Mr Mitchell advanced the submission that these changes were in any event immaterial. He referred me to the decision of the Court of Appeal in Raiffeisen Zentralbank Osterreich AG v. Crossseas Shipping Ltd [2001] 1 WLR 1135, which he said was authority for the proposition that in the case of written contracts an alteration to the contract after signature did not invalidate it unless it was material in the sense of being “potentially prejudicial to the legal rights or obligations of the affected party”: that, he said, was plainly not the case here.

I do not accept those submissions.

In the first place, I am not sure that the evidence establishes that Mr Grisay or the other clients implicitly authorised (or ratified) the change in the identity of the strip as between the draft and final versions of the Option Agreement and Sale and Purchase Agreement. Although the letters of 25 November 2003 do indeed refer in terms to the 2004 stock, there is nothing in KBPB’s letter which specifically draws clients’ attention to the fact that this is a change or that it will require alterations to the documentation which they had already signed: some may have noticed that, and recognised the implications, but others may not.

But even if I were wrong about that, in Koenigsblatt v. Sweet the document which was altered after signature was the same document as that which the party had originally signed. I have been referred to no authority which deals with the situation in the present case - that is, the taking of a signature page from one document and its recycling for use in another. Mr Mitchell submits that there is no essential difference between the two situations. It should, he said, make no difference whether the (ex hypothesi, authorised) alterations are made to the selfsame document or whether, as is increasingly easy with modern technology, they are incorporated in a tidier form in a reconstituted document and the signature page from the earlier version is reattached. I do not agree. The parties in the present case must be taken to have regarded signature as an essential element in the effectiveness of the documents: that is to be inferred from their form. In such a case I believe that the common understanding is that the document to be signed exists as a discrete physical entity (whether in a single version or in a series of counterparts) at the moment of signing. The significance of this is not entirely talismanic (though it would not affect my view even if it were): the requirement that a party sign an actual existing authoritative version of the contractual document gives some, albeit not total, protection against fraud or mistake.

Further, even if I were wrong about the legitimacy of transferring signature pages in general, there is the additional factor that each of the three key documents in the present case was intended to be a deed. S. 1 (3) of the Law of Property (Miscellaneous Provisions) Act 1989 reads as follows:

“An instrument is validly executed as a deed by an individual if, and only if –

(a)

it is signed -

(i)

by him in the presence of a witness who attests the signature; or

(ii)

at his direction and in his presence and the presence of two witnesses who each attest the signature; and

(b)

it is delivered as a deed by him or a person authorised to do so on his behalf.”

Mr Bird submitted, and I agree, that that language necessarily involves that the signature and attestation must form part of the same physical document (the “it”) referred to at (a) which constitutes the deed. Mr Mitchell observed that, although these documents were expressed to be deeds, it was not necessary that they should be. I am not sure that that is correct, at least in the case of the Option Agreement, for which no consideration is given; but, even if it were, the fact remains that the parties intended them to be deeds and their validity must be judged on that basis.

Mr Mitchell placed considerable reliance on para. 20 of Mr Masters’ second witness statement in these proceedings, which reads as follows:

“I have already made it absolutely clear in my first witness statement … that draft documents were signed by the investors ahead of all other parties signing, including the final signatory, the bank. This is perfectly normal business practice. In modern commercial times it is not practical for multi-party contracts, deeds and other instruments to be signed in the same place at the same time.”

I cannot see how this assists his case. HMRC’s objection to the course followed in the present case was not that the documents were signed on different occasions but that the document signed by the clients did not correspond to the final version. The difficulties referred to by Mr Masters - which do not in fact seem to me to be specific to “modern commercial times” -can be and very commonly are met by deeds, or other contracts, being signed by different parties on different occasions or by the use of counterparts. There is no evidence - nor, for what it is worth, am I aware - of any general practice of signature pages being detached from an incomplete draft and attached to a later, and significantly different, version.

As for Mr Mitchell’s fall-back argument based on materiality, I cannot accept that the change from reference to 2003 stock to 2004 stock was immaterial. No doubt from a commercial point of view it will have been a matter of indifference to the client which stock was used. Nevertheless the transactions entered into for the purpose of the Scheme were intended to be real contracts creating real legal rights and obligations. From that point of view, the identity of the stock which was the subject matter of those rights and obligations was necessarily not only material but fundamental. Raiffeisen does not assist the Claimants. It is concerned with the operation of a particular, penal, rule affecting the enforceability of contractual instruments such as guarantees. It is not concerned with the separate question of whether a party who has undertaken certain contractual obligations in relation to subject-matter A can be treated as bound by equivalent obligations in relation to subject-matter B.

Accordingly, I cannot accept the Claimants’ primary case that the substitution of the signature pages had no effect on the validity of the contracts in question or, thus, of the Scheme as a whole. I accept that the flaws on which HMRC rely are essentially formal. But I see nothing wrong in applying a strict test of formality to the validity of the agreements with which we are concerned in this case. Their entire raison d’être is to create - and demonstrably to create - a series of formal legal relationships: if they do not do that, they do nothing.

Although Mr. Mitchell proceeded on the basis that the error apparent on the face of the “chronology” which I describe at para. 17 above was being said to represent a distinct flaw in the implementation of the Scheme, I do not in fact believe that to be correct. As I note at para. 23 above, all that Mr Preston appears to be saying in para. 2.4.10 of the Information is that the erroneous placing of step 12 is most easily explained on the basis that what was envisaged was signature of “a draft document” and that that supported his suspicion that clients had been asked to sign draft documents in advance of completion of the final versions. I am not sure how strong that reasoning really is; but what matters for present purposes is that HMRC appear to rely on the point only as supporting evidence of the primary fraud and not as a distinct fraud in its own right. (Having said that, there is some basis for Mr Mitchell’s understanding to the contrary: I discuss this at para. 51 below.)

That being so, I think that the chronology may be something of a red herring. But I should say that I cannot accept Mr Mitchell’s submission that HMRC should have attached no importance to it because they should have appreciated that it was a KB-generated document. If, contrary to what I understand to be the case, the Court was being invited to suspect that the plan was that the client would definitively execute the final version of the Option Agreement prior to acquiring the option I do not believe that HMRC were bound to treat that simply as an error by KB: it would be entirely reasonable to assume that the steps to be followed were carefully planned between KB and Mercury.

Mr. Mitchell argued that there was no inconsistency between the contents of the minute and the terms of the later e-mail correspondence; but I was unpersuaded. It seems to me that there was good reason to suspect that the minute was created after the event and purported to record an agreement which had not at that date been reached. His other point was that by the date in question the putative tax loss had already accrued, so that even if the authors of the minute – on the face of it KB and not Mercury – had misrepresented the position that affected only the subsequent “unwinding” and not the tax claim itself. That is true as far as it goes. But it does not mean that the procedures adopted can cast no light on what occurred at an earlier stage of the Scheme.

I believe that the Information shows reasonable ground to suspect that there were serious flaws in the implementation of the Scheme. I turn to the separate question of whether there was reasonable ground to suspect that those flaws were the result of dishonesty on the part of the Claimants.

WAS IT REASONABLE TO SUSPECT DISHONESTY ?

Whether there were reasonable grounds to suspect dishonesty is, in principle, an objective question, depending on the assessment of the Court and not the view of HMRC. But it nevertheless requires an exercise of judgment and evaluation. I am inevitably in a better position to carry out that evaluation than Judge Spencer was because I have had the advantage of inter partes argument and considerably fuller evidence than was before him; but it is important that I remind myself that I am not in a position analogous to that of a Judge on the return date following the grant of ex parte interim relief. Under the 1970 Act the decision whether to issue a warrant can only be taken once for all by the Judge to whom the application under s. 20C is made. Subject to one important qualification, my role is thus in principle confined to one of judicial review: I have to ask not what I myself would have decided on the material which I have seen but whether the Judge could properly have come to the conclusion that he did on the material before him. The qualification relates to disclosure: it is well-established that, even if the Judge’s decision is in itself unimpeachable, I can quash the warrants if HMRC misled by the Court either by positive mis-statements or by failing to draw its attention to material matters.

The first question is whether Judge Spencer’s decision – or, therefore, the decisions of Judge Hillen and of the Board – can be impugned on the basis of the material in that information. The question for him is whether “there [was] reasonable ground to suspect” dishonesty. The only authority to which I was referred on the effect of that language was O’Hara v. Chief Constable of the Royal Ulster Constabulary [1997] AC 286. That confirms that the test is objective, but beyond that it is of limited assistance. It seems to me that as a matter of principle it is important in assessing the reasonableness of any grounds advanced for suspecting fraud not to set the bar too low. The judgment of what grounds for suspicion it is reasonable to require must be sensitive to the context. As has been repeatedly emphasised since the decision of the House of Lords in Inland Revenue Commissioners v Rossminster Ltd [1980] AC 952, the powers under s.20C are highly intrusive, and solid grounds for suspicion should be required before their exercise can be justified. I express the view at para. 68 below that the grounds advanced in the present case – at least in the light of all the material before me – are not particularly strong. But, having said all that, I accept Mr. Bird’s submission that “reasonable ground to suspect” appears to set a somewhat lower threshold than “reasonable ground to believe”; and I cannot conscientiously say that it was irrational of the Judge to conclude that the test was satisfied in this case, particularly if – as he was led to believe – there was evidence that Mercury had concealed from its own clients the change in the stock to be used.

I turn therefore to the question of disclosure. As I have said, HMRC were under a duty not only to avoid positive misrepresentation but “to make a full disclosure of all matters which might affect the court's decision to make or refuse the order and, in particular, to make disclosure of all matters known to [them] which might militate against the making of an order” (see per Bingham LJ in R. v. Crown Court at Lewes (1991) 93 Cr. App. R. 60, at p. 68). The Claimants at para. 49 of their Amended Grounds, which did duty also as Mr Mitchell’s skeleton argument, identified nine alleged breaches by HMRC of their duty to give full and frank disclosure. But Mr Mitchell also made two more general points on the way in which the case was presented to Judge Spencer with which it is convenient to deal first.

First, Mr Mitchell submitted that the way in which HMRC formulated its case of dishonesty was inconsistent and liable to give a misleading impression. In particular, the terms of the summary at section 1 – set out at para. 18 above – do not correspond well with the detailed exposition of the suspected fraud in section 2. Section 1 describes the suspected fraudulent implementation of the Scheme as having been “envisaged from the outset” (albeit only “in part”): see para. 1.4. That suggests a particularly serious and cynical dishonest intent on the part of the Claimants. But the principal fraud suspected – that is, the substitution of the signature pages – is said in section 2 to have been resorted to only when the timetable became too tight: it could hardly, therefore, have been envisaged from the start. Para. 1.4 appears in fact to have been intended as a reference only to the erroneous sequence provided for in the “chronology”, which is accorded in section 2 a very minor place in HMRC’s reasoning (see para. 23 above); but that is not apparent from the way in which it is expressed. Likewise, para. 1.6 in section 1 refers – though, again, this is not apparent, to a very different kind of fraud from the principal fraud. I have some sympathy with this criticism. Of course it is in the nature of a summary that it cannot cover all the subtleties that would appear in a complete exposition; but in my view the introductory section of the Information does lead the reader to expect a rather different kind of fraud from that which is developed in section 2. However it does not seem that the Judge was misled. It is clear from his observations to Mr Preston that he – correctly – understood the essence of the suspected fraud as lying in the “ringing” of the key documents (a term borrowed from motor vehicle fraud) and that that was said to be a response to problems with the timetable which had not originally been anticipated.

Secondly, Mr. Mitchell complained that the Judge had not been shown any documents. I am bound to say that I find it surprising that the Judge was being asked to reach a conclusion about a suspected fraud whose essence depended entirely on the effect of a number of documents without being supplied with copies of the documents in question. In an application for ex parte relief in the civil courts, copies of relevant documents would be supplied as a matter of course. It may be that the practice in the criminal courts, deriving from the traditional format of an information, has been different; but conventional practice should not be unthinkingly followed when it gets in the way of giving proper assistance to the court. The authorities repeatedly emphasise that the approval of a judge to the issue of search warrants – which unless properly justified represent a gross intrusion on civil liberties - cannot be a rubber stamp: it is his duty to subject the information put before him to jealous scrutiny. Such scrutiny will be more difficult, and may sometimes be impossible, if relevant documents are sought to be summarised (however accurately) rather than being supplied to the Court. However, whether the non-production of copy documents involves a breach of the duty of disclosure must depend on the particular case. In the present case I do not think that it would be sufficient by itself to justify the quashing of the warrants. Whether or not sight of actual copy documents would have helped the Judge to understand the allegation about the recycling of the signature pages (which is debatable), it is not in issue that that occurred. And I find it hard to see how it could have assisted him in judging whether that procedure may have been dishonest. There is one particular respect in which he might conceivably have been assisted by sight of a copy document, but the resulting error was not material: see paras. 61-62 below.

I turn to the particular misrepresentations/non-disclosures pleaded by the Claimants.

Mr. Masters’ Character

The Claimants say that the Court should have been told not only – as it was – that Mr. Masters was a solicitor but that “he had an unblemished practising record and was of good character”. I do not believe that that was a fact of sufficient materiality to require to be drawn to the Court’s attention. It could at most be of marginal significance in deciding whether there was reasonable ground to suspect dishonesty: the Judge was not being asked to make any kind of final adjudication.

The Claimants say that the Judge should have been told that Mercury had co-operated fully in the investigation and that HMRC had expressed themselves satisfied with the documents provided: see para. 7 above. Again, I do not regard that as a matter which required to be expressly spelt out. Mr Mitchell drew attention to a reference in the Information to documents having been provided “piecemeal” and said that that implied a lack of co-operation; but I do not agree. Mr Mitchell also made the point in his oral submissions that the key documents which were being relied on by HMRC – including the draft agreements signed by Mr Grisay - had in fact been supplied by Mercury, which he said was strongly suggestive that they did not believe that they had anything to hide. But Mr Bird made clear that HMRC were not convinced that the draft agreements had been supplied intentionally – the equivalent documents from Mr Clayton and Mr Meadon had not been supplied – and that they were concerned that Mercury had never volunteered the fact that it had adopted the procedure of transposing signature pages. Those suspicions, which are not articulated in the Information, may or may not be justified; but the fact that they were held illustrates that the question of co-operation was not necessarily straightforward and that it was fair to leave it alone.

As appears from the passage from the Information set out at para. 24 above, the Judge was given the clear impression that there was reason to suppose that Mercury’s clients had not been informed of the change from 2003 to 2004 stock prior to the conclusion of the agreements. That was said to follow not only from the re-use of the signature pages but from the fact that there was no trace in Mr Kowalishin’s papers of his having been told of the change prior to the crucial date. If that was correct, it would support the suspicion that the Claimants had acted dishonestly, because it would suggest a desire to conceal the change in the position from the clients and, therefore, a motive for deliberately taking short cuts with the documents even though they knew that it was wrong.

Mr Mitchell submitted that the letters of 25 November 2003 which I have set out at para. 35 above, to which the Judge was not referred, entirely undermine that argument. Even if it was not specifically drawn to clients’ attention that the use of the 2004 strip represented a change, the fact that they were expressly asked to sign a letter referring to it is inconsistent with any intention to conceal from them that the 2004 strip was to be used.

In my judgment that submission is well-founded. The terms of para. 2.4.22 of the Information, and the failure to refer to the letters of 25 November 2003, constitute an important misrepresentation/non-disclosure. In my judgment that non-disclosure was material. That conclusion seems to me inescapable in view of the Judge’s expressed understanding that the misleading of the clients was at the heart of the fraud. In the exchange set out at para. 27 above he put it to Mr Preston that a scheme using 2004 as opposed to 2003 stock was “not the deal which the scheme users thought they were getting”; to which Mr Preston replied “that’s right”. I do not believe that the Judge’s question, or Mr Preston’s answer, would have been possible had the letters of 25 November been before the Court.

I will defer consideration of the consequences of this misrepresentation/non-disclosure until I have considered the remainder of the matters relied on by the Claimants.

Mis-Stating the Law

The non-disclosure relied on under this head is framed as follows:

“Wrongly invited the Court to draw inferences of ‘fraudulent alteration’ of documents which they knew, or ought to have known, were unsound in law”.

This is not really an allegation of non-disclosure. In any event it falls away in view of my conclusions at paras. 38-42 above.

Attributing the “Chronology” to Mercury

Mr Preston said at para. 2.4.10 of the Information (see para. 23 above) that he suspected that the chronology was produced by Mercury. Mr Mitchell submitted that there were no grounds for that suspicion and that in fact it was fairly clear that it was produced by KB. I agree; and, as I have observed, the Judge was not in a position to make his own assessment because no copy of the document was produced to him.

The question however is whether that was a material mistake. In my judgment it was not. The error in the chronology was deployed essentially only in order to support a conclusion which, as it turns out, is not in issue – namely that Mercury always intended to proceed by obtaining signatures to draft documents which would later be transferred to the final versions. In any event, it is entirely reasonable to assume that Mercury and KB co-operated closely on the steps to be taken and accordingly that Mercury would have been aware of the contents of this document (which was indeed disclosed to HMRC during the investigation as part of Mr Clayton’s file).

The Summary of the Minute of 4 December 2003

The non-disclosure relied on under this head is framed as follows:

“Wrongly asserted that the minute of the 4^th December 2003 records an offer by Schroders for the purchase of gilt strips/options to buy gilt strips from Olivos and then says that in fact Schroders had no contact with Olivos. In fact that minute does not refer to Olivos at all.”

I find that rather hard to penetrate, and what is alleged could not be properly analysed without setting out in full the terms of para. 2.4.26 of the Information, which I have summarised at para. 15 above. However, I do not regard that as necessary. I am satisfied that the Information fairly identified the essential points about the Minute, namely its reference to events which it was reasonable to suspect had not at that point occurred.

“Normal Commercial Practice”

The Claimants plead that HMRC should have told the Judge “that it is perfectly normal commercial practice for a party to sign a multi-party deed in advance of its final execution”. But that of course does not address the flaw in the implementation of the Scheme relied on by HMRC.

The position therefore is that I have found that the Court was misled in one, though only one, respect material to its conclusion that it was reasonable to suspect dishonesty on the part of the Claimants. I was not referred to any authority which directly addresses the question of the proper approach to deciding what consequences should follow in the case of non-disclosure in an application under s. 20C. In two cases concerning restraint orders, the Court of Appeal has emphasised that the position is not analogous to that obtaining in civil proceedings, because of the additional element of the public interest. That point was made expressly by Longmore LJ in J v. Crown Prosecution Service [2005] 4 All ER 391 ([2005] EWCA Civ 746), at para. 64 (see also per Laws LJ at paras. 54-57), and by Hughes LJ in Director of the Serious Fraud Office v A [2007] EWCA Crim 1927, at para. 18. In the latter case Hughes LJ said:

“Whilst it is appropriate to insist on strict compliance with the rule of disclosure, discharge of the order does not necessarily follow as a means of disciplining the applicant, at least absent what Longmore LJ in Jennings [i.e. J] referred to as 'so appalling a failure' that that ultimate sanction should be applied.”

The gist of the judgments in both cases is that the Court needs to exercise a broad discretion which recognises the importance of supervising this highly intrusive jurisdiction but also has regard to substantial justice and the public interest. So far as the public interest is concerned, it seems to me that there may be a difference between applications for restraint orders and for search warrants. In the case of the former there will typically be a continuing need for the order to remain in place, whereas in the latter case the warrant will have been executed once and for all and the principal consequence of its being quashed will be to open the way to a claim for damages. But that seems to me simply one factor to take into account: it does not call for a wholly different approach.

Adopting that approach, I have come to the conclusion that the decision of the Court to issue the warrants was unlawful. My reasons are as follows.

First, although the suggestion that clients were not informed that 2004 stock would be used was not, as a matter of logic, a necessary part of the grounds for suspicion which HMRC advanced, it did in fact form an important element of the case as presented both in the Information and in the oral submissions. As I have observed, it plainly weighed with the Judge. I am certainly not in a position to say that he would have reached the same decision in the absence of this element, or even that it is likely that he would have done so.

Secondly, this does not seem to me to have been a case where the grounds to suspect dishonesty on the part of the Claimants were particularly strong. It is, to put it no higher, entirely plausible that they took the course of substituting the signature pages simply as a convenient shortcut, without appreciating that it might invalidate the Scheme. If they really believed that that was its effect, but decided to proceed nonetheless in the hope that HMRC would never find out, they would be taking a very big risk for very little advantage: the essence of such schemes (assuming that they work in principle) is to get the paperwork formally right. The explanation offered by Mr Preston in the course of his oral submissions to the Judge - i.e. that the Claimants might have been too embarrassed to go back to their clients for a further set of signatures - seems to me rather flimsy. The errors in the chronology and the minute of the meeting of 4 December 2003 do not seem to me to shed much light on the Claimants’ state of mind in making the entirely separate decision to effect the substitution of the signature pages. In short, if I had now to decide, with the benefit of the fuller picture available to me, whether search warrants should be issued I do not believe that I would find that sufficiently good grounds had been shown.

Thirdly, although I have only found one clearly material misrepresentation/non-disclosure, I cannot say that this was a case where the presentation of the application otherwise was entirely faultless. There was another, though immaterial, error in the description of the documents (see paras. 61-62 above); the summary of the case in section 1 of the Information was not very coherent (see para. 51); and I think it was unfortunate that none of the key documents were exhibited to the Information.

Fourthly, I can see no serious damage to the public interest if these warrants are quashed. The only possible damage that I can see would be if the return of any documents seized had the effect of precluding a prosecution that could and should otherwise be advanced. I will hear argument on this aspect if required, but as at present advised this does not seem to me a likely outcome.

Overall, this seems to me to have been something of a borderline case for the deployment of the nuclear weapon of an application under s. 20C. If HMRC were to persuade the Court that this was a proper case for the issue of search warrants (covering not only business premises but the private homes of over twenty individuals) on an ex parte basis, it was incumbent on them to put their case with scrupulous accuracy and in such a way that the Judge was able to make a fair assessment of the grounds for suspicion being put forward. That did not occur.

The case having been decided specifically on the basis of non-disclosure to the Court, in my provisional view the only decisions that fall to be quashed are those of Judge Spencer and Judge Hillen. I will however hear submissions on that question and on any other questions as to the precise form of relief to be granted.