UT/2024/000069 - [2025] UKUT 00236 (TCC)
Upper Tribunal Tax and Chancery Chamber

UT/2024/000069 - [2025] UKUT 00236 (TCC)

Fecha: 23-May-2025

Domestic authorities

Domestic authorities

46.

The case of Impact, like the appeals before us, involved the use of MUCs. However, there is a critical distinction: the appellant (“ICSL”) in Impact was an intermediary whose customers were temporary work agencies and whose suppliers were the MUCs. The issues in Impact were focused, therefore, on whether HMRC could deregister ICSL where it had not itself fraudulently evaded VAT but had facilitated the VAT fraud of another and whether it mattered if ISCL was making supplies unconnected to the fraud and whether a de-registration in those circumstances would breach EU principles of proportionality, fiscal neutrality or legal certainty.

47.

Accordingly, the focus in Impact was not on the perpetrator of the VAT fraud but on those persons who could be said to be facilitating it by entering into supplies “connected” to the VAT fraud.

48.

Having set out the CJEU case law authorities (most relevantly for our purposes the cases of Halifax, Kittel, Ablessio and Cityland), Falk LJ noted at [55] that, if ISCL was correct, then, even if a taxable person was party to a conspiracy to defraud the revenue, HMRC would not be able to de-register the party if the actual VAT default was committed by another party to the conspiracy. She went on to note at [56] that:

“VAT fraud has proved to be a very significant problem. If the tools available to HMRC were limited in the way that ICSL maintains that they are, then they would be likely to be of limited effect in preventing future abuse. Importantly, there would be nothing to prevent the relevant person’s participation in further fraudulent schemes, with HMRC attempting to play “catch up”, trying to close the proverbial stable door after the horse has bolted. In contrast, deregistration is prospective in effect and, as is obvious, will prevent a trader from using its VAT number in fraudulent transaction chains in the future.”

49.

Falk LJ held at [58] that the lawfulness of HMRC’s decision to de-register ISCL depended on two different matters: (1) whether ISCL knew, or should have known, that it was taking part in transactions connected with the fraudulent evasion of VAT; and (2) whether deregistration was, in the particular circumstances of ICSL, proportionate.

50.

Falk LJ considered at [62] that it was of significance that, in both Ablessio and Cityland, references to fraud were included in the context of the broader Halifax abuse principle:

“… Properly understood and in the context of that case law, I consider that those references are intended to make it clear that the decisions in those cases do not prevent tax authorities taking steps to counter VAT fraud, including in relation to VAT registration. However, and as HMRC accept, those steps must be proportionate.”

51.

Falk LJ then emphasised, consistent with the overarching nature of the principle and its central importance in the fight against VAT fraud, the width of the abuse principle:

“63.

… The reference to both misuse of a VAT number and “other VAT fraud” in Ablessio at [38] also indicates that the CJEU was not intending to be prescriptive in describing the particular type of fraud that was required. (The reference at [30] to fictitious activity “in particular” was rightly not relied on by Mr Margolin as limiting what was said only to activity of that nature.)

64.

It is true that the reference in Ablessio at [36] to “the taxable person’s fraudulent intentions” is arguably more consistent with ICSL’s case. The same can be said of the reference to “tax fraud committed by that taxable person” in Cityland at [47]. However, I do not consider that, by those brief references, the CJEU was intending to limit the scope of the principle in that way ...

65.

Further, even read as ICSL submits, the references in these paragraphs are far from definitive. Ablessio refers at [36] only to a “suspicion” of fraudulent intentions as a factor to take into account, and as already indicated other references to “misuse” are broader. In Cityland, the preceding paragraph, [46], refers more broadly to a “risk to tax revenue and a likelihood of VAT fraud”, and at [47] the immediate context is the stated need to examine the taxable person’s conduct to ascertain the “nature and extent” of “any” tax fraud committed by them, so as to assess whether deregistration is an “appropriate penalty for ensuring the collection of all the VAT and combating VAT fraud”. It does not state in terms that they must have fraudulently evaded VAT.

66.

The dispositif in Cityland refers to a requirement for the tax authority to analyse “the nature of the infringements committed and the conduct of the taxable person at issue”, which is considerably broader than ICSL’s case would suggest.”

52.

Falk LJ dealt with the requirement for proportionality in these terms:

“67.

What Ablessio and Cityland do emphasise is the requirement to comply with the principle of proportionality. Deregistration cannot be based on mere suspicion. Rather, there must be “sound evidence giving objective grounds for considering that it is probable that the VAT identification number assigned to that taxable person will be used fraudulently”, and the decision must be based on an “overall assessment” (Ablessio at [34]). The “nature and the degree of seriousness of the infringements committed” must be examined (Cityland at [45]).

….

72… I cannot see a logical basis to distinguish between those who evade or may evade VAT themselves and facilitators as a matter of principle. Of course, the proximity and extent of a facilitator’s involvement in VAT fraud are likely to be relevant factors in determining whether the tax authority’s action is proportionate on the facts, alongside other factors including the extent of the anticipated untainted supplies, but that is very different to ruling out action in relation to VAT registration altogether.”