Add a bookmark to get started

Aerial view of abstract staircase
11 January 202417 minute read

Tax controversy in 2023: has inflation of taxing provisions been tamed?

This article was originally published in Tax Journal, [15 December] and is reproduced with permission from the publisher.

Have HMRC been pushing their interpretation of tax legislation too far? Jason Collins and Ravikaran Ahlawat (DLA Piper) look at lessons from some of this year’s key judgments.

HMRC has a deliberate strategy of losing cases. A former permanent secretary said that if HMRC didn’t lose some cases, it would not be doing its job properly, as it would not be establishing the boundaries of the law. HMRC boasts a success rate of at least 80%, rising to 90% in recent years. Is 2023 the year in which this win percentage started dropping back, and will 2024 also see this win ‘inflation’ tamed further?

Many taxpayers and practitioners will recognise the move by HMRC towards stretching existing legislation to the max. It is not a surprise that HMRC has done this – for years the courts grappled with how far they could go with ‘purposive construction’ of tax legislation. A low water mark for HMRC was Mayes [2011] EWCA CIV 407, where the Court of Appeal (CA) found that gaps in legislation which was highly prescriptive could not be ‘filled in’ by the courts, and the Supreme Court (SC) refused HMRC permission to appeal. The high water mark is probably the SC’s decisions in UBS [2016] UKSC 13 where the court surprised everyone and took a very expansive view of the root of the legislation to find against the taxpayer. Since then, numerous cases have flown through the courts with judges at all levels very happy to trounce the taxpayer. This has led to the battle-weary joke that the purpose of tax legislation is to raise tax, so the courts will always find for HMRC – and has led to a tremendous amount of uncertainty.

But is there a sense that HMRC, and tax authorities in other jurisdictions, are taking this push too far? Judging by a recent article in the Financial Times (‘Global tax clampdown fuels boom in insurance against costly rulings’, 26 November 2023), US insurance underwriters are taking this view, with a growing appetite for insuring tax risks, based on the view that fewer tax assessments will prevail as tax authorities are over-reaching. Are we seeing that in the UK? Let’s have a look at some of the ‘avoidance’ cases in 2023 where HMRC hasn’t been victorious.


HMRC defeats in ‘avoidance’ cases


Delinian Ltd (formerly Euromoney Institutional Investment PLC) v HMRC [2023] EWCA Civ 1281 concerned share for share exchange tax deferral. The taxpayer entered into an agreement to sell its shares in a company, the consideration being c75% settled with an exchange of shares in the acquirer and c25% in cash. The head of tax at the taxpayer became involved and analysed that tax on the gain would be deferred in respect of the exchange of shares under the rules in TCGA 1992 ss 135–137, but chargeable in respect of the cash element. He asked if the deal could be restructured so that, instead of cash, the acquirer issued redeemable preference shares in the same amount. Those shares would also benefit from the deferral in ss 135–137 and, if held for more than one year, the redemption for cash would benefit from the substantial shareholders exemption (SSE) and thus be tax free. The acquirer agreed to the change and the transaction proceeded on that basis.

HMRC used this evidence of the head of tax’s involvement in changing part of the terms of the deal to argue that there should be no tax deferral for any of the exchanges – as the exchanges needed to have been effected for a bona fide commercial purpose and not form part of a scheme or arrangements of which a (or the) main purpose is to avoid a liability to tax. HMRC lost before the FTT and UT. The CA also rejected HMRC’s case. The CA rejected the idea that ‘scheme or arrangements’ was anything other than the whole transaction, and the ‘exchange’ anything other than all of the shares exchanged. It held that HMRC’s approach of sifting through evidence to find exchanges/arrangements within those exchanges/arrangements was wrong. The FTT had made findings of fact about the relative importance of the tax saving, and that the transaction would have proceeded even if the acquirer had not agreed to the preference shares. HMRC had not challenged those findings of fact and as such, on the CA’s approach to the legislative test, HMRC’s appeal failed.

The taxpayer had also argued that using the SSE intentionally afforded to it by Parliament was not ‘avoidance’ for the purposes of ss 135–137. The CA disagreed, holding that the right to defer tax under ss135–137 is very clearly lost if it is used for an ulterior (main) purpose to avoid tax in the future, even if that later avoidance is permitted.


In Fisher and another [2023] UKSC 44, HMRC failed to persuade the SC to attribute the income of an offshore entity to a UK resident individual under the ‘transfer of assets abroad’ anti-avoidance legislation. A company had transferred its business to a Gibraltar company, as many other had done in the gambling industry when a new form of betting duty was introduced. HMRC sought to charge the shareholders in that company on the profits arising to the Gibraltar entity. The SC held that the transfer of assets must have been by the individuals for the provision to be engaged, and that a transfer by a company cannot be treated as a transfer made by its shareholders. The court preferred a narrower interpretation of the legislation to provide certainty in its application. It also commented that, if the legislation leaves a gap, it is not for judges to fill through an expansive approach to interpretation; new legislation is required, with all the checks and balances that that entails. Hopefully other courts will take heed of this approach.

GE Financial

GE Financial Investments v HMRC [2023] UKUT 146 (TCC) concerned claims for a double tax treaty credit. A UK company was liable to tax in the US on the basis that its stock was ‘stapled’ to a US incorporated company (i.e. its shares could only be transferred in lockstep with a transfer of the US company shares) and as such it was deemed to be a US corporation for most purposes of the US tax code. Its US profits arose through participation in a US partnership. As a UK incorporated company, it was also liable to tax on those profits in the UK and claimed a double tax treaty credit for the US tax. HMRC argued it was not entitled to the credit because it was not liable to tax in the US based on being ‘resident’.

Article 4(1) of the treaty stipulates that ‘residence’ for treaty purposes encompasses anyone liable to tax by reason of domicile, residence, citizenship, place of management, place of incorporation, or other similar criteria.

The FTT held that stapling was not a ‘similar criteria’ to any of the others listed, particularly residence, because stapling involved no physical nexus for the stapled company with the country concerned. The UT overturned this decision. It held that treaties must be interpreted purposively and not by the literal meaning of their words. The UT held that the purpose of article 4(1) is to determine when a country imposes ‘full’ (i.e. worldwide) taxation, and the list was to identify common methods of doing so – so that any method which achieves this same object is a ‘similar criteria’ to those listed. The US tax code applies full taxation on a company incorporated in the US and treats a stapled foreign company for most intents and purposes as if it were a US one. Deemed incorporation is sufficiently similar to incorporation.

HMRC took this case because the UK company was part of arrangements to route income into the UK, to give the UK group more capacity for interest deductions before hitting the corporate interest restriction, without that routed UK income generating any additional cash tax in the UK (due to the tax credit). HMRC saw this as part of an attempt to erode the UK tax base, as it now sees so many instances of debt structuring (see, for example, the line of cases on unallowable purpose). Others might just see this as prudent tax planning.


HMRC testing the limits of legislation

2023 saw some interesting defeats for HMRC at the limits of certain legislation, where HMRC is not suggesting that a tax avoidance motive is in play. We have picked out two here.

SSE Generation

The dispute in the capital allowances case of HMRC v SSE Generation Ltd [2023] UKSC 17 was about whether certain assets in a hydroelectric plant were ‘tunnels’ or ‘aqueducts’ as defined in CAA 2001 s 22(1)(a) List B. If classified as such, they would not qualify for capital allowances. The taxpayer succeeded. The SC affirmed the application of the ‘immediate context’ rule, meaning that while words in statutes generally take their ordinary meanings, they can be shaped or limited by their surrounding words and the overall context in which they appear. This is particularly so where a word can have more than one ordinary meaning – and in trying to establish Parliament’s intention, trying to find a theme linking words is permissible. In this case, that theme was assets that relate to transportation. A unique feature of the hydroelectric plant is that many of the assets are underground and so what might be tunnels and aqueducts in one ordinary sense of the word, were not tunnels and aqueducts falling within such a transportation scheme. The SC agreed with the CA in the outcome, albeit that it differed in how far it would restrict the meaning of aqueduct – preferring any bridge-like structure for carrying water, over the CA’s approach of such a structure for carrying only a canal.

Gap Group

Gap Group Ltd v HMRC [2023] UKFTT 970 (TC) is a VAT case where the dispute was whether there was a single or separate supplies. The taxpayer supplied equipment to construction and engineering customers. Equipment was hired out with a full tank of red diesel, and customers were charged for refuelling if the equipment was returned with less than a full tank. Gap maintained that the fuel was a separate supply, subject to a reduced VAT rate, in contrast to the standard rate applied to the hire charges. HMRC, however, contended that the fuel formed an integral part of the plant hire service, and as such the taxpayer was carrying out a single supply at the standard VAT rate.

The FTT found there to be a separate supply. It recognised the customers’ autonomy in choosing to refuel the equipment and the clear separation of charges for hire and fuel. The supply of fuel was merely a convenience offered at the end of the hire period, not enhancing the hire but forming a standalone choice.


Procedure and practice

2023 also saw some interesting cases around the administration of taxes.

Parker Hannifin

In Parker Hannifin (GB) Ltd v HMRC [2023] UKFTT 971 (TC), HMRC was enquiring into whether the purposes for which this company was party to a loan relationship included an ‘unallowable purpose’ under CTA 2009 s 441/s 442. It issued a formal information request dictating the search terms that should be applied to emailboxes of those involved in setting up the loan relationship. The taxpayer did not appeal the notice. It instructed PwC to carry out the search. PwC reported to HMRC that c11,000 documents responded to the search terms but less than 2,000 were relevant to the enquiry and provided those to HMRC. HMRC wanted to see the rest.

The taxpayer then appealed the notice to the FTT. It was granted permission to do so out of time. By the time the appeal came to be heard by the FTT, HMRC had accepted that, based on PwC’s explanation of how it had ruled documents out, around 4,000 were irrelevant. But it still wanted to see the rest.

The FTT said that if the notice had been appealed prior to any compliance with it, it would have set it aside as being too broad. But it had to determine what was before it, namely whether in the context of the disclosure made in response to the notice, the notice should be varied to mirror what had been provided, varied in some other way or left as is.

HMRC did not challenge the competency with which PwC had carried out its review. HMRC’s case was that the unseen documents would be informative and provide context. One suspects that the truth of the matter is it wanted to determine for itself whether something was relevant. The FTT made it clear that HMRC’s powers only allow it to see documents which are relevant – so it was not in its power to see ‘source’ documents and carry out its own sift. The FTT varied the notice to add a qualification that the notice was to produce documents responding to the search terms ‘which relate to [the purposes of the company]’.

In practice, we would rarely advise a client to accept an information request based just on search terms – but in any event the ratio of this case would suggest that any information request (even if it describes a class of documents rather than specifies search terms) is always subject to the implicit qualification that only relevant documents need to be produced. In contrast to civil information powers, HMRC’s criminal powers expressly allow them to remove potentially irrelevant material to be reviewed later. It is also proper for the taxpayer to self-evaluate relevance; there is nothing unusual about this in civil litigation and if the other side has reason to believe relevant material has not been disclosed, both systems include mechanisms for probing further.


Refinitiv Ltd (and others) v HMRC [2023] UKUT 257 (TCC) concerned the extent of an advance pricing agreement (APA). In that case, within a corporate group, various UK companies provided services to an associated company in Switzerland to help that company enhance valuable intellectual property that it held. HMRC entered into an APA to the effect that the arm’s length price for the UK services would be calculated on a ‘cost-plus’ basis for periods up to 2014. The APA was not renewed and, in a turn for the worst HMRC, for the periods from 2015 to 2018, charged diverted profits tax. HMRC argued that the ‘relevant alternative provision’ that would have existed but for tax considerations was enhancement services between two independent parties being charged on a more valuable ‘profit split’ arm’s length basis. In respect of 2018 specifically, Switzerland had disposed of the IP for a gain and the profit split thus included a share of that gain as well as a share in the income from exploiting that IP in the year up to the disposal.

The UK companies sought a judicial review for 2018 on the basis that HMRC was bound to recognise that it had agreed to a cost-plus for the services provided up to 2014, so it could not now tax the element of the gain which represented enhancement arising during that period, as the UK had already been rewarded for those services. By way of analogy, HMRC’s position is a little like if a person were to carry out building services at your house for agreed charges, but then claim to be entitled to a share of the increase in the value of the house arising from those services.

Refinitiv UK unsurprisingly failed in the judicial review claim – because, quite simply, as the APA did not expressly cover 2018, HMRC was not acting contrary to the agreed terms and there was therefore nothing in public law to stop HMRC from doing what it pleased for 2018. However, that is not the end of the road. The UK companies still have a right of appeal to the FTT about whether HMRC’s approach of applying the profit split without regard for the reward HMRC was happy to agree was correct up until 2014 is correct. The appeal will no doubt raise interesting points about whether the real-world arm’s length provision for 2018 can take account of prices agreed purely for the purposes of a tax computation for earlier periods (i.e. on a fictitious basis).

Although obiter dictum, the UT said HMRC would not have been bound by the APA even if the issue was using a profit split to charge extra profits for corporation tax rather than DPT for 2018. Two points can be taken from this case: first, the judicial review was never likely to succeed, and the attempt to land a procedural ‘knockout blow’ was the sort of procedural skirmish we’ve previously cautioned against (‘Tax disputes in 2021: beware the procedural skirmish’ (J Collins & L Redhead), Tax Journal, 10 December). The second is that when agreeing an APA for service provision to enhance an asset, seek to add a term that HMRC will observe the agreed basis for pricing in any future chargeable period in which the asset might be disposed of. If HMRC agrees to such a term, that should be enough to ensure the APA relates to that later, albeit as yet unknown, period and that HMRC is bound by it.

Separately, the taxpayers established legal professional privilege over various documents HMRC had requested under a Sch 36 para 35 notice (Refinitiv and others v HMRC [2023] UKFTT 222 (TC)). The tribunal judge reviewed the withheld documents and was satisfied that legal advice, but not litigation, privilege applied.


G Ives v HMRC [2023] UKFTT 968 (TC) considers the scope of the FTT’s jurisdiction. The taxpayer bought and sold residential properties over a number of years, treated each as a capital transaction and claimed principal private residence (PPR) relief in respect of the gains. HMRC issued a closure notice stating he had omitted ‘property developing income’ on the basis that he was engaged in a trade, but HMRC had also warned in prior correspondence that its alternative case was that they were capital disposals for which PPR relief would not be available. The taxpayer argued that this alternative position was not within the jurisdiction of the FTT to hear as it did not form part of the closure notice.

It has become relatively settled law that HMRC is not precluded from arguing a new point to support an existing conclusion in a closure notice. The taxpayer argued that his was a different case, as HMRC was arguing for an alternative conclusion and a different tax (CGT not income tax).

An appeal lies to the tribunal in respect of the ‘matter in question’. Although admitting to be vexed by the fact that the alternative involved a different tax, the FTT concluded that the matter in question was the officer’s conclusions, and that could include the views stated in the earlier correspondence linked to the closure notice. Perhaps there might have been a different result had HMRC formed this view after the closure notice had been issued.


Final thoughts

Whilst this article contains a one-sided view of taxpayer wins, we believe the cases cited do point to a trend in the judiciary towards a more restrictive view of anti-avoidance rules and failed attempts by HMRC to test the limits of tax legislation. 2024 will be an interesting year of cases, including the hearings by the CA of the three test cases under the ‘unallowable purpose’ interest deductibility rules.

The authors thank Lauren Redhead, senior associate at DLA Piper, for her contribution to this article.