DECISION
1.
Following a hearing lasting some 14 days before Cyril Shaw and myself in
October 2008, we released a decision in which we decided, on the basis of our
understanding of the CJEU's judgement in Axel Kittel v Belgium 2006 ECR I-6161 (“Kittel”), that S&I was not entitled to credit for its input
VAT on a number of its purchases of mobile phones in deals in which it had
purchased and exported them between April and June 2006.
2.
We reached this conclusion because we concluded that those deals were
connected with fraud and that, although we were not convinced that S&I knew
that these transactions were connected with fraud, it should have known that
such was the case.
3.
In deciding whether S&I should have known of the connection we
applied the following test, namely whether a reasonable man with ordinary
competence in the position of S & I, and knowing what S & I knew (a)
would have taken any additional steps, and (b) would have come to the
conclusion, on the basis of what he knew and had found out, that it was more
likely than not that the transaction was connected to fraud.
4.
Both parties appealed against the decision. The Upper Tribunal's
decision on the appeals is to be found at [2012] STC 1620. After our decision was
released, and before the hearing before the Upper Tribunal, the Court of Appeal
decided the appeal in Mobilx Ltd (in administration) v Revenue and Customs
Commissioners and other actions [2010] STC 1436. The test of knowledge we
had applied was there held to have been wrong. Moses LJ said:
“[59] The test in Kittel is simple and should not
be over refined. It embraces not only those who know of the connection but
those who "should have known". Thus it includes those who should have
known from the circumstances which surrounded their transactions that they were
connected to fraudulent evasion. If a trader should have known that the only
reasonable explanation for the transaction in which he was involved was that it
was connected with fraud and if it turns out that the transactions was
connected with fraudulent evasion of VAT then he should have known of that fact.
He may properly be regarded as a participant for the reasons explained in Kittel.
“[60] The true position to be derived from Kittel
does not extend to circumstances in which the taxable person should have known
that by his purchase it was more likely than not that his transaction was
connected with fraudulent evasion. But a trader may be regarded as a
participant where he should have known that the only reasonable explanation for
the circumstances in which his purchase took place was that it was a transaction
connected with such fraudulent evasion."
.
5.
Thus, instead of asking ourselves whether it was more likely than not
that the transactions were connected to fraud we should, as the Upper Tribunal
pointed out, have asked whether:
"[46] " [S & I] knew or should have known
that [its] the transactions were connected with fraud or that there was no reasonable
possibility other than they were connected with fraud?
6.
The Upper Tribunal was not able to conclude on the basis of the findings
we had made in our decision whether or not this test was satisfied and remitted
the case back to the First-tier tribunal. It was directed that this tribunal
should consider the matter on the basis only of the evidence which had been
deployed in considering the appeal in 2008.
7.
By the time that of remission Mr. Shaw had retired. The appeal was therefore
remitted to me sitting alone. Thus I have to decide whether on the evidence
before the tribunal S&I should have known that its transactions were
connected to fraud.
8.
I should recall at this stage that we considered 90 deals and found that
in 79 of them it was shown to our satisfaction that there was a connection to
the fraudulent evasion of VAT (that is to say that both a connection to the
alleged defaulter and its default were proved); in the remaining eleven we
found that HMRC had not proved that there was such a connection or default, not
that there was no connection or default. The Upper Tribunal did not disapprove
of these findings.
9.
Mr. Patchett-Joyce made a number of submissions on the proper scope of
the second limb of the Kittel knowledge test in particular in the light
of the recent CJEU judgements in (a) the joined cases of Mahageben kft v
Nemetzi Ado-es Vamhivatal Del-dunantuli Regionalis Ado Foigazgatosage C-80/11
(“Mahageben”) and Peter David v Nemetzi Ado-es Vamhivatal Eszak-alfoldi
Regionalis Ado Foigazgatosage C-142/11(“David”) and (b) Toth v Nemetzi
Ado-es Vamhivatal Eszak-magyarorszagi RegionalisAdo Foigazgatosage C-324/11 (“Toth”).
These are discussed below, and for the reasons I shall explain, I reject them.
However if it subsequently transpires that I am wrong, it seems to me that
where I can provide a coherent formulation of an alternative test on the basis
of those submissions, time could be saved if I offered my conclusions in
relation to those tests. That I have attempted to do.
10.
In summary: Mahageben related to the supply of logs which RK had
invoiced to M. M had used the logs to make its own supplies [43] but the tax
authorities thought that RK could not have supplied them. They refused the
deduction on the grounds that they had no delivery evidence and that M had not
acted with due diligence as the Hungarian law required. David: Mr David
had invoiced for labour he had supplied. The labour had not been his own and he
had been invoiced for its supply to him by X. X could not substantiate to the
tax authorities what he had provided; Mr David had also supplied labour to a
contractor and had used a subcontractor which had invoiced him. But there are
doubts about the subcontractor who may not have fulfilled its tax obligations.
The question before the court was what conditions could be imposed on the right
to deduct income tax. Toth: Mr Toth had undertaken building work using
subcontractors including ML who had not complied with his tax obligations or
declared his employees who as a result worked in the black economy. The CJEU
addressed whether ML's tax fraud prevented Mr Toth’s deduction - holding that
it did so only if the Kittel principle applied.
Mr Patchett-Joyce’s submissions
(1). Would Have Had to have Known.
11.
Mr. Patchett-Joyce says that the proper question is not whether the
taxpayer "should have known" but whether it "would have had to
have known" of the relevant fraud. He says that this is the correct
translation of the original French of the judgement in Kittel.
12.
To my mind however, unless "would have had to have known"
means the same as "must have known" which I think means in effect the
same as “knew”, I can see no difference between “should have known” (or “ought
to have known”) and Mr Patchett-Joyce’s preferred phrase. That is because: the
second part of the test must mean something different from the first; it is
plainly an objective test and in that context “would have had to have known”
conveys to me no higher standard of objective intellectual compulsion than
“ought to have known”.
13.
If, contrary to my belief, there is only one limb to the knowledge test,
namely whether one can conclude that the trader knew, then the tribunal found
that it was not proved that it did.
(2). The relevant connection is limited to the fraud of
S&I's supplier, and the requisite knowledge is limited to knowledge of the
fraud by its immediate supplier.
14.
The issue of the nature of the requisite connection has not been
remitted to me by the Upper Tribunal and Mr. Patchett-Joyce recognises that
that question may only be pursued elsewhere. But he says that the question of
the requisite knowledge is before me, and effectively that even if the relevant
connection may be to a fraud by a person before the immediate supplier in the
chain of supply, the Kittel principle applies only if the trader had the
requisite knowledge of a fraud by his supplier.
15.
In this connection Mr. Patchett-Joyce points to the following passages
in Mahageben and Toth as showing that the requisite knowledge is limited
to knowledge of connection to the immediate supplier’s fraud. In para [45] of Mahagaven
the court states the relevant test by reference to "transaction" in
the singular: "knew, or ought to have known, that transaction was
connected with fraud previously committed by the supplier or another trader at
an earlier stage in the transaction." The emphasis on “the transaction” in
the singular, Mr. Patchett-Joyce says, points to the transaction in which the
trader was involved, not to earlier transactions in the chain. That emphasis he
says is repeated in M where at paragraph [66] the court provides a caveat
saying that "the taxable person is not in possession of any material
justifying the suspicion that irregularities or fraud have been committed
within that invoice issuer’s sphere of activity." That he says points to
the supply to the trader not to an earlier transaction in the chain.
16.
I do not share Mr. Patchett-Joyce 's view of these cases.
17.
In David the trader was denied input tax inter alia because it had
not acted with the due diligence required by Hungarian law. The question was
whether a condition so to act could be imposed by the State. The court, having
explained that the question was in the context of the assumption that the
transaction had been carried out and that a proper invoice had been supplied
([44]), said that in such circumstances the "only" way in which input
VAT could be denied was on Kittel grounds:
[45] In those circumstances, a taxable person can be refused
the benefit of the right to deduct only on the basis of the case law resulting
from paragraphs 56 to 61 of [Kittel], according to which it must be
established, on the basis of objective facts, that taxable persons to whom were
supplied the goods or services which served as the basis on which to
substantiate the right to deduct, knew, or ought to have known, that the
transaction was connected with fraud previously committed by the supplier or
another trader at an earlier stage in the transaction."
18.
By contrast the court said ([47]) that if a trader did have the
requisite knowledge a system of strict liability (such as a condition for due
diligence) went beyond what was necessary:
"47. By contrast, it is incompatible with the rules
governing the right to deduct under that directive, ... to impose a penalty, in
the form of refusing that right to a taxable person who did not know, and
could not have known, that the transaction was concerned was connected with
fraud committed by the supplier or that another transaction forming part of the
chain of supply prior or subsequent to that transaction carried out by the
taxable person was vitiated by VAT fraud ...”
“[48]. The establishment of a system of strict liability
would go beyond what is necessary to preserve the public exchequer's rights
...".
19.
In these paragraphs the court is not limiting or refining the Kittel
test but saying that because that test is the only way input VAT can be
denied to an otherwise compliant transaction, the due diligence requirement
went too far. Thus at [49] it says:
"49. Given that the refusal of the right to deduct
in accordance with paragraph 45 of the present judgement is an exception to the
application of the fundamental principle constituted by that right, it is for
the tax authority to establish, to the requisite legal standard, the objective
evidence which allows the conclusion to be drawn that the taxable person knew,
or ought to have known that the transaction relied on as the basis for the
right to deduct was connected with fraud committed by the supplier or by
another trader acting earlier in the chain of this of supply."
20.
This was not a case concerned with the precise formulation of the Kittel
test and the court does not address that issue. The most that can be said is
that the Court sets out its understanding of that test. That it does in
slightly different ways in paragraphs 45, 47 and 49: in paragraph 45 seemingly
by reference only to "the transaction" in the singular, but in
paragraph 47 by reference to transactions forming part of the chain of supply
prior or subsequent to the trader's transaction, and in paragraphs 49 and 52 by
reference to a transaction carried out by another trader acting earlier in the
chain of supply. These slightly varying formulations do not to my mind indicate
an understanding of the Kittel test as requiring knowledge of fraud only
in the immediately preceding transaction: the clear later references to another
transaction in the chain of supply make clear to me that, at most, the concentration
on "the transaction" in the language of paragraph 45 is a reference
to the fact that in the case under consideration the potential fraud related
only to the provision of the subcontracted services to the trader.
21.
The same in my view is true of the Court's analysis in David
where the dispositive paragraph [66] excepts from the State’s obligation not to
impose due diligence requirements on the trader as a condition of input tax
recovery circumstances where the trader "is in possession of material
justifying the suspicion that irregularities or fraud have been committed within
that invoice issuer's sphere of activity". Here the court was addressing
the question whether the right to deduct could be refused where the taxpayer
did not satisfy himself of matters related relating to the bona fides of his
supplier ([51]). In that context [52], quoting its earlier reasoning in David,
it said that the right to deduct could be refused only where the Kittel
conditions were satisfied and then went on to consider the ability of the State
to impose additional obligations ([55]). The measures the State could impose
depended on the circumstances ([59]) and if there were indications of fraud
([60]) a trader could be required by the State to make further enquiries as a
condition of getting his input VAT, but "as a general rule" ([61])
the State could not impose blanket due diligence requirements. Thus the
formulation in the disposition "suspicion [of fraud] within the invoice
issuer's sphere of activity" is not a recasting or refinement of Kittel,
but the definition of a possible exception to the general rule precluding the
imposition of due diligence tests.
22.
Thus to my mind in neither its decision in David nor that in Mahageben
did the CJEU intend to refine, limit or explain the knowledge test in Kittel,
nor did its description of that test provide any new insight into it.
23.
In my view Kittel, Mahageben and David indicate
clearly that the relevant question is whether the trader knew or ought to have known
(or should have known) that its transaction was connected with fraud by someone
in the chain of supply.
24.
Toth concerned a trader who had been refused an input tax
deduction for the invoiced supply of the services of a subcontractor. The
subcontractor had defaulted on certain tax obligations and there were
irregularities in its VAT compliance. Mr. Toth had not investigated the
relationship between the subcontractor and the workers. Four questions were
referred to the CJEU of which numbers 2 to 4 were discussed before me.
25.
Question no 2 was whether the fact that the issuer of the invoice
employed workers in the black economy could prevent Mr. Toth from being
entitled to deduct income tax. The court answered that question thus:
"[39]. Accordingly, the answer to the second
question is that [the directive] must be interpreted as meaning that it
precludes the tax authority from refusing a taxable person the right to deduct
VAT due or paid for services provided to him on the ground that the issuer of
the invoice relating to those services did not declare the workers he employed,
without that authority establishing, on the basis of objective evidence, that
the taxable person concerned knew, or ought to have known, that the transaction
relied on as a basis for the right to deduct was connected with fraud committed
by the issuer of the invoice or by another trader acting earlier in the chain
of supply.
26.
I see nothing in that answer which suggests any limitation on the
relevant knowledge for the purpose of the Kittel test.
27.
Question no 3 was whether the directive must be interpreted as meaning
that the fact that a taxable person had not verified whether a legal
relationship existed between the workers employed at the work site and the
issuer of the invoice, or whether the issuer had declared those workers
constitutes "an objective factor which demonstrates that the addressee of
the invoice, knew or ought to have known that he was participating in a
transaction involving fraudulent evasion of VAT”.
28.
The court referred to its conclusion at para [66] of Mahageben
quoted above. It said that it also applied to the supply of services:
"[44] That conclusion in relation to a supply of
goods also applies in the case of a supply of services as regards the question
whether it may be considered that the taxable person knew or should have known
that the supply relied on for entitlement to his rights to deduct was connected
with fraud committed by the issuer of the invoice, on the grounds that he had
not verified whether the issuer of the invoice had the necessary employees
available to be able to supply the services at issue, whether that issuer had
fulfilled his obligations as regards declaration of those employees and whether
the employees of the issuer and carried out the work at issue.
"[45] Consequently, the answer to the third question
is that [the directive] must be interpreted as meaning that the fact that a
taxable person did not verify whether either a legal relationship existed
between the workers employed on a work site and the issuer of the invoice or whether
the latter had demonstrated those workers does not constitute an objective
factor which demonstrates that the addressee of the invoice knew ought to have
known that he was participating in a transaction involving fraudulent evasion
of VAT, where the addressee was not in possession of any material justifying
the suspicion that irregularities or fraud had been committed within that
invoice issuer's sphere of activity. Accordingly the right to deduct may not be
refused on that ground where the material and formal conditions laid down by
that directive for the exercise of that right are met."
29.
Mr. Patchett-Joyce describes [45] as a worked example of the application
of the Kittel test. The test the court there applies, he says, is to
look solely at the knowledge of fraud within the invoice issuer's sphere of
activity.
30.
I accept that the court is applying the Kittel test, but it is
applying it in the context of the narrow question of whether the simple failure
to verify demonstrated knowledge, and against the factual background in which
the only possible fraud was in the invoice issuer's activities. The court had
earlier, at [39], recognised the breadth of the Kittel test when it used
the words "acting earlier in the chain of supply", but it was then
applying that test in a situation where it was alleged that the issuer of the
invoice had committed the fraud. These paragraphs do not indicate that the
court thought that test of relevant knowledge should be limited to knowledge of
the supplier's fraud.
31.
I also note that a suspicion of fraud might affect the relevant
conclusion.
32.
Question no 4 was whether all the circumstances of the issuer and its
activities could be taken into account in assessing whether the supply had in
fact taken place. The court answered that question thus:
“[53]. Having regard to the foregoing considerations the
answer to the fourth question is that, where the tax authority provides
specific evidence of the existence of fraud, [the directive] and the principle
of tax neutrality do not preclude the National Court from verifying on the
basis of an overall examination of the circumstances of the case, whether the
issuer of the invoice carried out the transaction in question himself. However,
in a situation such as that at issue in the main proceedings, the right to
deduct may be refused only where it is established by the tax authority, on the
basis of objective evidence, that the addressee of the invoice knew or should
have known that the transaction relied on as a basis for the right to deduct
was connected with a fraud committed by the issuer or another operator
supplying inputs in the chain of supply."
33.
Mr. Patchett-Joyce calls attention to the words "or another
operator supplying inputs in the chain of supply”. He says that the
italicised words must be there for a purpose and that purpose is to limit the
principle to persons providing something to the taxpayer - the "chain of
supply" was not the entire commercial chain (the sale of clay to the brick
maker, the sale of the bricks to the wholesaler, their sale to the builder, and
the builder’s supply of building services), but the triangular supply in which
labour was provided by X which was supplied to the trader by Z, or logs were
conveyed by X and their supply invoiced by Y (to whom X supplied labour). They
limited the required knowledge to knowledge of fraud by a supplier to the
taxpayer.
34.
I do not agree. I see nothing in the judgement which suggests such a
limitation. It seems to me that "another trader acting earlier in the
chain of supply" and "another operator supplying inputs in the chain of
supply" are the same thing.
35.
Further the sentence in which these words appear is prefaced by "in
a situation such as that at issue in the main proceedings". In those
proceedings the only concern was over fraud at the level of those providing the
services used by Mr. Toth.
36.
Thus I find nothing in Toth which illuminates the nature of the
relevant knowledge for the purposes of the Kittel test.
37.
If I am wrong however and test were to be found to be relevant knowledge
of the trader's immediate suppliers' fraud then I would conclude that the
evidence before the tribunal did not persuade me that either S&I either
knew or should have no such fraud nor that there was such a fraud.
(3) Outside the scope of VAT.
38.
At [36] in Mobilx, Moses LJ said:
"The court’s reference in paragraph 55 to a quartet
of previous decisions reinforces the proposition that fraudulent tax evasion
falls outwith the scope of VAT and thus the scope of the right to deduct input
tax. Fraudulent evasion of tax does not meet the objective criteria, such as
whether the activity is "economic activity" or a taxable person is
"acting as such", by which the scope of VAT and of the right to
deduct are identified.”
39.
Mr. Patchett-Joyce says that if the result of the application of this
principle is such that a transaction is outside the scope of VAT then in
relation to that transaction there cannot have been any VAT chargeable on it
and any attempted evasion of VAT which would have been due had it not been
outside the scope cannot be the fraudulent evasion of VAT. Thus any form of
knowledge of a fraudster’s activity would not be knowledge of VAT fraud for the
purposes of the Kittel test.
40.
It seems to me that this is a problem only if one accepts the premise
that falling within the Kittel provisions takes a transaction outside
the scope of VAT. But that is not my understanding of principle.
41.
The Kittel principle concerns the right to deduct. It does not
address the tax chargeable on the supply which would normally give rise to the
corresponding right of deduction. In Kittel the court said that
objective criteria for the concept of supply and economic activity are not met
where tax is evaded by a the taxable person himself ([53]), and it applies that
principle to deny input credits to a person who knew he was participating in
fraudulent evasion ([54-59]). To my mind the court cannot have intended that what
it described as participation in the fraudulent evasion by a person seeking to
deduct was not such participation because of the fraud in which he was
participating. The initial evasion must therefore be treated as evasion of VAT
for the purposes of the later test whether or not that evasion would cause that
initial transaction to fall foul of the principle at paragraph [53].
42.
I therefore conclude that this argument does not affect the question of
the existence of the relevant knowledge which has been remitted to me.
(4) The reasonable businessman
43.
In addressing whether S&I should have known of a fraud in our
original decision we applied the test of what the reaction of a reasonable
businessman would have been to the facts known to S&I at the relevant time.
44.
Mr. Patchett-Joyce criticises this use of the reasonable businessman in
determining what the taxpayer should have known. He says that (i) its use was
criticised by Lewison J in Livewire [2009] STC 643at [122] to [125], and
(ii) the test in Kittel does not require the concept or consideration of
what a reasonable businessman would have done, it simply requires the
commissioners to prove, having regard to objective factors, that the taxpayer
would have had to have known that its transactions were connected with fraud.
45.
In Livewire the tribunal had considered what facts should be
treated as known by the taxpayer company and what skill it should be treated as
having. It considered a test in the Insolvency Act 1986 in relation to the
question it set itself of determining experience of a reasonable businessman and
concluded that the taxpayer should not be treated as endowed with the knowledge
and skill of a competent director but:
“The test that we apply is
accordingly whether a person with the knowledge, skill and experience of the
director concerned would have known that the transactions were connected with
fraud.”
46.
Lewison J said:
“[123] It is common ground that the supposed analogy with
section 214 of the Insolvency Act 1986 is at best unhelpful and at worst
positively misleading. First, section 214 requires the court to take into
account both (a) the general knowledge, skill and experience that may
reasonably be expected of a person carrying out the same functions as are
carried out by that director in relation to the company and also (b) the
general knowledge, skill and experience that that director has. In other words
(b) is knowledge over and above the minimum to be expected of an ordinarily
competent director. It does not allow a lower standard to be adopted. Second,
the Kittel test applies to the taxable person. The taxable person was Olympia (the company). The question therefore for the Tribunal was not what a director of Olympia knew or ought to have known, but what the company itself knew or ought to have
known. The knowledge of a director of the company may, to be sure, be
attributed to a company, but there may be other knowledge (for example that of
a senior employee) which, on the facts ought also to be attributed to the
company: Meridian Global Funds Management Asia Ltd v Securities Commission
[1985] AC 500. Accordingly, in applying the test of what ought to have been known
by a director with the knowledge, skill and experience of the particular
director concerned the Tribunal, in my judgment, fell into a legal error. To
the extent that a domestic analogy is appropriate, the Tribunal applied a lower
standard than that which would have been appropriate to support a finding of
constructive knowledge. …
“[124] These errors did feed into the Tribunal's ultimate
conclusions. The director on whom the Tribunal concentrated was Mr Habib.
Having dealt with a number of the precautions that Olympia did take the
Tribunal concluded (§ 56):
"Weighing up all these factors, we consider that Mr
Habib was, on account of his inexperience, naïve and gullible. Had we decided
that the test of "ought to have known" should be based on an ordinarily
competent director we might well have decided that such a person ought at least
to have made further enquiries, but we have based the test on the experience of
the particular director, Mr Habib….."
“[125] Quite apart from the difficulty of imagining a naïve
and gullible company, the Tribunal expressly adopted a legal test that
required fewer precautions (or a lower level of understanding) than would have
been required of a director of ordinary competence. The Tribunal does not
positively find that it would have decided that an ordinarily competent
director would have made further enquiries, but in my judgment the fact that
the Tribunal considered that it "might well" have done so indicates
that the application of the correct legal test might alter the outcome. In
applying the correct legal test, the Tribunal must consider not only the
knowledge that should be attributed to the company via its directors, but also
the knowledge that should be attributed to the company via its senior
employees. In this context, knowledge includes both knowledge of facts and the
ability to evaluate those facts and to draw appropriate conclusions from them.”
47.
My reading of the remarks of Lewison J in the quoted paragraphs does not
coincide with that of Mr. Patchett-Joyce. It seems clear to me that he is not
criticising the use of the reasonable businessman test but the particular
standard imposed by the tribunal in that case.
48.
In relation to Mr. Patchett-Joyce's second criticism, I believe it
evaporates on further consideration. The second limb of the Kittel test
requires the tribunal to determine whether a taxpayer should have known of the
fraud. That must be an objective test, and it must be made by reference to
objective factors. But, having set out the objective factors, how is one to
determine whether they lead to a conclusion that a person should have known?
The process of reasoning from fact to conclusion is a human one and it requires
a human to do it. That raises the questions of what sort of human: a child, an
overzealous customs officer, or a reasonable man, and of what degree of
knowledge to attribute to him. (Whatever else, it seems to me that it cannot be
a question of what the appellant actually concluded because that is the same as
knowledge.) The only human entity in a position to conduct this exercise is the
tribunal, and to my mind use by the tribunal of the "reasonable
businessman" in its decision-making is merely an attempt to describe the
mindset it adopted in taking that decision. By using "reasonable" it
indicates that it did not attempt to clothe itself in the mindset of a child, a
paranoid customs officer or habitual VAT villain, but retained its own
(presumed reasonable) mindset; and by the use of "businessman"
described itself as having some knowledge of commercial transactions. It is
thus a description of the deducting mindset of the tribunal in approaching the
facts (and a denial of any pretence of being something different) rather than
the creation of a new test. Thus in this decision it would make no difference
if I replaced "the reasonable businessman" with "I".
(5) Can it be found that a trader should have known of
fraud when it is not proved that fraud does not exist in the chain or where it
is shown that there is no fraud?
49.
I have noted above that we found that it was not proved that in eleven
of the 90 deals there was a connection to a fraud. The question arises in the
circumstances of those eleven deals as to whether it is logically possible to
find at the same time: both that the taxpayer should have known that they were
connected to fraud, and that is not proved that they were. That question is
acute because the objective factors which convinced us that a reasonable
business man would have concluded that it was more likely than not that the
transactions were connected to fraud were the same in every case. There was no
difference between those factors in relation to a deal not shown to be linked
to fraud and one which was.
50.
This question is more disturbing when one is applying the test in Mobilx.
Applying that test, the should have known limb is satisfied if the only
reasonable explanation of the facts is that the taxpayer’s transactions are
connected with fraud: but if the only reasonable explanation of what was known
to S&I was fraud, how can the tribunal find that it was not proved that
they were connected with fraud - particularly where the tribunal will have had
details of the sales and purchases in the chain of supply which were not
available to S&I?
51.
Put another way, does our original finding that some deals were not
proved to be connected to fraud necessarily entail a conclusion, not only that
S&I could not have concluded that the only reasonable explanation of the
circumstances of those eleven deals was fraud, but also, because the information
available to S & I in all the other cases was the same, that in no case
could S & I have so concluded?
52.
Mr. Davis-White answered this thus. He gave the example of a person so severely
injured in a car crash that a bystander might say that the only reasonable expectation
was that he would die, but unbeknown to him a skilled surgeon was on hand who
actually saves him. The "only reasonable explanation" allows for the
possibility of another explanation. Further Mr. Davies-White says that the
conundrum disappears when the possibility of the trader taking further steps
(making further enquiries) is considered.
53.
Mr Davis-White also points to para [59] of Moses LJ’s judgement in Mobilx
where he says:
“If a trader should have known that the only reasonable
explanation for the transaction in which he was involved was that it was
connected with fraud and if it turns out that the transactions was
connected with fraudulent evasion of VAT then he should have known of that
fact.”
54.
Mr Davis-White says that the italicised words show that Moses LJ
recognised that one could reach a conclusion that the only reasonable
explanation of a transaction was connected to fraud when the transaction was
not in fact proved to have been so connected.
55.
I am generally with Mr. Davis-White on the these points. The only
reasonable explanation does not preclude the existence of unreasonable but
possible explanations. However in the eleven deals in which we found that the
connection was on balance not proven, a conclusion that the only reasonable
explanation was connection to fraud might call into question our finding that
the deals were not proven to be connected. That finding in each case was
reached on the evidence of (i) the chain of supply to S&I and (ii) the
alleged fraud of a person in that chain: but our decision betrays that we did
not consider the factors relevant to S&I’s knowledge in reaching our
conclusions as to whether those deals were connected with fraud. That may be
therefore a fault in our decision as regards those details, but it is not one
which may be corrected by me in this decision. But because we concentrated on
the direct evidence in relation to the fraud and the chain of supply rather
than the wider circumstances, and because we did not find that there was not
fraud, it leaves open to me to find that the only reasonable explanation of the
circumstances of the deals in these cases was that they were linked to fraud.
56.
Had we found that it was proved that certain deals were not connected to
fraud, in my view, one could only then find that the only reasonable
explanation for the other deals was connection to fraud if either the
circumstances of those deals were outside the reasonably possible or if there
was evidence distinguishing their circumstances from the others.
57.
(7) Measures required: what further steps could S&I have been
reasonably expected (or assumed) to have undertaken for the purpose of
determining what it should have known?
58.
Mr. Patchett-Joyce raises the question as to whether the objective
factors on which the decision is to be made are limited to those actually known
to the trader, or whether there should be added to the mix those facts which
the "reasonable businessman" would have found as a result of any
investigation he would have felt compelled to make. In my judgement such facts
are to be added to the mix for the reasons in the following paragraphs. But I
state my conclusions in S&I’s case on the basis (a) that I am correct in
this judgement, and (b) that I am not.
59.
To my mind the formulation of the test, using as it does the words
"should", "ought", "would have had to", raises
the question of what conclusion to reach if you have a suspicion. The answer
depends on the strength (and evidence for) the suspicion. If what you see
seriously worries to you then you should, ought or would have had to,
investigate further; if it is a mere passing fancy you would not. Thus what you
should, ought or would have had to know must include what you would have found
out. I find some comfort in this conclusion in the caveat of the CJEU in Toth
where at paragraph [45] it says that the lack of due diligence is not an
objective factor demonstrating knowledge where there was no
"suspicion" of fraud. Where there was such suspicion, the failure to
investigate could therefore possibly be an objective factor. That could only be
because it suggests that either the failure to investigate demonstrates
knowledge, or that such investigation should have been undertaken and would
have revealed fraud. Likewise in Mahageben at [60] reference is made to
a reasonable trader:
“ It is true that where there are indications pointing to
infringements or fraud, a reasonable trader would, depending on the
circumstances of the case, be obliged to make enquiries about another trader
from whom he intends to purchase goods or services in order to assess the
latter’s trustworthiness ...".
60.
This is in the context of a potential fraud by a supplier but the
acceptance of the obligation to make enquiries is described in the context of a
reasonable trader.
61.
In Kittel the court held that a person who did not know or did
not have "the means of knowing" of fraud was not barred from the
right of deduction; the converse was the trader who "knew or should have
known" of the fraud. In Mobilx Moses LJ said [51] that the CJEU
must have intended the phrase "know or should have known ... to have the
same meaning as "knowing or having any means of knowing". Against
that background it is clear to me that in determining whether the only
reasonable explanation of the relevant circumstances was fraud, one must assume
that the trader has knowledge of that which he has the means of discovering and
which he should or ought to discover.
62.
In our decision we concluded that S & I’s knowledge of the
circumstances surrounding its transactions was such that it would have given
rise to a "very serious concern, and possibly a conclusion that it was
more likely than not, that each of S&I’s April, May and June transactions
would have been connected with fraud" [214]. We continued at [215] to say
that a reasonable businessman in these circumstances would have undertaken
further investigations and that these circumstances and those investigations
and would have revealed the fraud.
63.
Mr. Patchett-Joyce says that Mahageben at [53] emphasises that:
"traders who take every precaution which could
reasonably be required of them to ensure that their transactions are not
connected with fraud ... must be able to rely on the legality of those
transactions without risk of losing their right to deduct ...".
He says that such precautions are limited to those which
are "required" of them, and that such precautions are properly
described as "measures":
"[59] In those circumstances, it follows from the
case law referred to in paragraphs 53 and 54 of the present judgement, that
determination of the measures which may, in a particular case reasonably be
required of a taxable person wishing to exercise the right to deduct VAT in
order to satisfy himself that his transactions are not connected with fraud
committed by a trader at an earlier stage of a transaction depends essentially
on the circumstances of that particular case.
“[60] It is true that, when there are indications
pointing to an infringement or fraud, a reasonable trader could, depending on
the circumstances of the case, be obliged to make enquiries about another
trader from whom he intends to purchase goods or services in order to ascertain
the latter’s trustworthiness."
64.
Mr. Patchett-Joyce says that the indications must relate to the
particular case - i.e. the particular supply: general awareness of a fraud will
not be sufficient to trigger an obligation; nor will simply the risk of
infringement. Further he says that this passage shows that the obligation is
"to make enquiries about another trader from whom he intends to purchase
goods or services" so that it is limited to the supplier, and is not more
extensive; and it is limited to the trustworthiness of his suppliers.
65.
I do not share Mr. Patchett-Joyce's interpretation of these passages. To
my mind they are directed to the specific question of when and whether a member
state may impose conditions on a taxpayer’s right to deduction. They are not
directly related to the application of Kittel test: that test does not
relate the conditions imposed by the state but whether the taxpayer has the
means of knowledge of connection to fraud. The “measures” to which the court
refers are not relevant to the test: in the context of evaluating whether
someone had means of knowledge, but measures which a state seeks to impose as
an additional condition for deduction.
66.
As a result I consider that in deciding what information was available
to a trader for the purpose of determining what he should have concluded, he
should be taken to have found out what a reasonable businessman in his
circumstances would have found out , and I do not see these passages as
proscribing such an approach.
67.
(8) Unistar Group Ltd, Unistar Trading Ltd v HMRC
68.
I should note that after the hearing had taken place I was kindly sent a
copy of the FTT’s decision in this case by The Khan Partnership. I read it but
it did not affect my thinking. I therefore did not seek HMRC’s comments on it.
Was the only reasonable explanation of S&I’s deals
that they were connected to fraud. And, if so, should S&I have known that?
69.
In our original decision we made a number of findings which are relevant
to this issue.
(i) A chain of transactions before S&I.
70.
S&I knew that there was a chain of transactions before each of its
acquisitions. It knew that its transactions were part of a back-to-back chain
of acquisitions and sales because it knew that traders did not want to hold
stock (see the quote from Mr. Ahmed at (vi)(a) on page 57). It was clear to me
that S&I was aware that contemporaneously with its purchase and sale there
was a series of virtually contemporaneous purchases and sales by UK traders.
(ii) Margins
71.
S&I were aware that it was a feature of the market that the margin
per phone made by “buffer” (UK to UK) trader in the chains were fairly standard
and less than that made by exporters.
(iii) Non UK Specification Phones
72.
The phones S&I dealt in were mostly of Central European
specification with plugs not intended for the UK market. Although it was
possible to change plugs relatively cheaply this could not be done in
back-to-back transactions. Thus S & I must have known that these phones
passed speedily through the hands of members of the chain without conversion
for the UK market.
73.
The only possible explanations that I can see for this are: (i) that the
phones had been converted in the UK (for why else were they in the UK) to non
UK specification prior to the start of the chain of contemporaneous
transactions; (ii) they were being bought and sold as commodities in a price
arbitrage market in which their precise specification did not matter; and (iii)
they were in the UK so that they could be exported.
74.
As regards (i) it seems to me that it is possible that UK specification telephones could be acquired in the UK and converted here to EU specification with
the intention of being sold on into the EU. It does not however seem reasonable
to suppose that such conversions would have been done on a speculative basis –
without the guarantee of a certain customer – on the kind of scale which
S&I’s transactions involved: whilst it seems perfectly reasonable that a
single consignment of phones might have been converted for a sale which failed
and which were then offered for sale more generally (or that a single
consignment of phones with European specification might have been brought into
the UK with the object of changing them ready for sale in the UK market, and
that perhaps circumstances changed and the holder received an offer he could
not refuse from a member of S&I’s chain so that they were then exported to
their “home” market by S&I) I can see no reasonable explanation of how this
could happen in such a large number of cases. I conclude that (i) is not a
reasonable explanation.
75.
So far as (ii) is concerned I can understand phones being dealt with as
commodities so that traders in the UK might buy and sell them even though they
were not configured for the UK market. But that does not explain why all the
consignments of phones were in the UK: such trades would not require the
movement of the phones across frontiers. Neither their export by S&I nor
their presence in the UK while being bought and sold by members of the chain
are consistent with that explanation. Indeed any transport is inconsistent with
such trade because it increases cost and diminishes margins.
76.
I conclude that the only reasonable explanation for the presence of
these phones in the UK was that they were brought into the UK for the purpose being exported after having been bought and sold by a chain of UK dealers who did not hold stock.
77.
(iv) Details on invoices etc
78.
The invoices and purchase orders sent and received for each of the deals
contained very little detailed description of the phones. That indicated to us
(and indicates to me) that neither S&I nor its customers were really
interested in those details. We found (and I find) that consistent with the use
of the phones for purposes which require their movement in a chain of
transactions involving their export but also consistent with their use as a
commodity in an arbitrage market. Those seem to me to be the only reasonable
explanations of that circumstance.
79.
But the second of these explanations is not reasonable in the light of
the other circumstances.
80.
In the case of non-UK specification phones arbitrage trading makes no
commercial sense for the reasons set out at [76] above Thus for these phones this
factor further confirms that the only reasonable explanation of the trade was connection
to some scheme involving their trading and export.
81.
So far as the UK specification phones were concerned, S&I’s
knowledge of the consistent margins in UK- UK trading is inconsistent with a
conclusion that these could be arbitrage transactions (because in such
transactions trading margins would not have been so uniform). Such a conclusion
would be unreasonable.
(iv) Knowledge of fraud
82.
Knowledge that the phones were being traded in the UK for the purpose only of exporting them has only one reasonable explanation namely that
they were connected with VAT fraud.
83.
S&I knew that there was significant VAT fraud in the market in which
it operated. It knew that some of its transactions had been connected to such
fraud. It also knew the nature of the fraud. So it knew that it involved the
default by one member of a supply chain in the payment of VAT and the export of
phones. It must have known that the only reasonable explanation of trade in the
UK for the purpose of exporting was VAT fraud.
84.
I conclude that S&I should have known that the only reasonable
explanation of the deals was that they were connected to VAT fraud. I reach
this conclusion without considering what additional information S&I could
have obtained.
85.
However even if there were a possible explanation of the other
circumstances of each of its deals other than connection to VAT fraud, it is
clear to me that the concerns detailed in our decision would have excited such
serious concern that they were linked to fraud that the only reasonable response
would have been to make the further investigations described in paragraph 215
of our decision, and that those investigations would have revealed that there
was a connection to fraud. That is because the information received would have
only one reasonable explanation, namely fraud. Thus if as I believe to be the
case, I am entitled to consider what S&I would have found out if it had
taken the steps a reasonable businessman in its position would have taken, it
would have concluded that its deals were connected to fraud.
Passages in our decision
86.
We addressed our decision to the test of likelihood of connection, and
provided our conclusions on the basis of the requirement that the test was met
if it was more likely than not. But there is in my view a link between what may
be called a risk based approach and the “no other reasonable explanation”
approach. It is this: an explanation that is highly unlikely is not a reasonable
explanation; one that is likely is reasonable. If the delivery van does not
arrive there may be a number of possible explanations. One is that the driver
has been eaten by a lion, another that she is stuck in traffic. The absence of
a zoo, a circus or reports of an escaped lion in the vicinity makes the first
of these an unlikely and therefore unreasonable explanation; the presence of a
nearby normally congested road makes the second a not unlikely and a reasonable
explanation; if in the circumstances that reason is more likely than not it
must logically be the most reasonable explanation. On that basis one might
characterise our decision as being that the most reasonable explanation was connection
to fraud.
87.
It remains to consider, against that background, whether any of our
earlier conclusions were inconsistent with those expressed here.
88.
At para [208], in relation to the deals with the identified suppliers,
we concluded that a reasonable businessman in S&I’s position would have
come to the conclusion that it was more likely than not that the deals were
connected to fraud. There is no inconsistency here.
89.
At para [201] we referred to the following factors which HMRC had said
pointed to uncommercial dealing: (i) counterparties acting variously as supplier
and purchase; (ii) back to back same day trades; (iii) small consistent margins
earned by buffers; (iv) higher margins earned by S&I; (v) no insurance;
(vi) limited description in invoices. Each gave us some measure of concern. We
concluded: “Taking these issues together we find they do not point
unequivocally to involvement in transaction[s] connected to fraud. But they
would raise in the mind of the reasonable businessman serious concerns about
such a connection.”. This remains my view. The facts would point
“unequivocally” to fraud only if there were no other possibility. In my
opinion S&I could have considered it theoretically possible that there was
no connection to fraud; but although that was a possibility it was not in my
view a reasonable one for the reasons I have explained above.
90.
At para 214 we said, in relation to the other deals that the factors we
noted would have given rise “to a very serious concern, and possibly a
conclusion that it was more likely than not that each of S&I’s...transactions
would have been connected to fraud.” It might be said that “possibly” was used
to indicate that there were other reasonably possible explanations. But that
fails to look at our conclusion in the round. We were sure (and I remain sure)
that the suspicions we identified would have justified very stringent further
investigation, and we were sure (and I remain sure) that that investigation
would have uncovered fraud. Because of the test we were applying had no need to
consider whether the identified issues on their own had no reasonable
explanation other than fraud. The words quoted merely indicate that we might
have been able to reach that conclusion earlier.
91.
My (and our) reaction to the deals was that the facts were such that if
we had been in the position of S&I, I (we) would have known they were
linked to fraud. We were persuaded that Mr. Ahmed himself did not reach that
conclusion, but we were clear that we would have done. At the time of writing
our decision we described that conviction against the test of coming to a conclusion
on balance, but, had the test been that, had we been S&I’s shoes we would have
concluded that the only explanation was fraud, we would have come to that
conclusion.
.Conclusion.
92.
I find that S&I should have known that its deals were connected to
fraud because that is the only reasonable explanation of the circumstances of
those deals.
Rights of Appeal.
This document contains full findings of fact and reasons
for the decision. Any party dissatisfied with this decision has a right to
apply for permission to appeal against it pursuant to Rule 39 of the Tribunal
Procedure (First-tier Tribunal) (Tax Chamber) Rules 2009. The application
must be received by this Tribunal not later than 56 days after this decision is
sent to that party. The parties are referred to “Guidance to accompany a
Decision from the First-tier Tribunal (Tax Chamber)” which accompanies and
forms part of this decision notice.
CHARLES HELLIER
TRIBUNAL JUDGE
RELEASE DATE: 5 April 2013
