William Reeves v The Commissioners for HMRC

This is an appeal against the decision of the First-tier Tribunal (Judge John Brooks) released on 28 February 2017 dismissing Mr Reeves’ appeal against HMRC’s disallowance of a claim under section 165 of the Taxation of Chargeable Gains Act 1992 (‘TCGA’) for holdover relief from capital gains tax ([2017] UKFTT 192 (TC)).

The issue in the appeal is whether Mr Reeves is entitled to claim holdover relief pursuant to section 165 TCGA on a disposal that he made when he gifted his interest in an LLP called BlueCrest to a UK resident company, WHR Ltd, of which he was the sole shareholder. BlueCrest LLP operated a hedge fund business. At the time of the disposal Mr Reeves was resident in the United States and non-resident in the UK for tax purposes. Accruing to Mr Reeves on this disposal was a capital gain of about £33.6 million. He asserts that that gain is held over and attaches to the interest now owned by WHR. If and when WHR disposes of it, WHR may have to pay that charge.

HMRC assert that Mr Reeves is not entitled to holdover relief because that relief is denied by the operation of section 167(2) TCGA. Mr Reeves accepts that the literal wording of the relevant provisions would mean that the disposal does not benefit from holdover relief. But Mr Reeves says that the words cannot possibly mean what they say because it leads to absurd and arbitrary results. He therefore invites us to interpret the words to prevent absurdity from arising. If they are so read, Mr Reeves does benefit from holdover relief.

Mr Reeves further contends that if the Tribunal is not able to construe the provisions restrictively using the ordinary canons of statutory construction then:

In an annex to this decision, we set out the text of the main relevant statutory provisions.

The facts are not in dispute and are set out in an agreed Statement of Facts and Issues. In summary:

Section 2(1) TCGA provides:

“2.

Persons and gains chargeable to capital gains tax, and allowable losses

(1)

Subject to any exceptions provided by this Act, and without prejudice to sections 10 and 276, a person shall be chargeable to capital gains tax in respect of chargeable gains accruing to him in a year of assessment during any part of which he is resident in the United Kingdom, or during which he is ordinarily resident in the United Kingdom.”

Thus the liability to capital gains tax depends generally on the residence or ordinary residence of the taxpayer being in the United Kingdom.

The taxation of limited liability partnerships is dealt with by section 59A TCGA, originally inserted by the Limited Liability Partnerships Act 2000, later substituted by the Finance Act 2001. That provides that where an LLP carries on a trade or business with a view to profit, the assets held by the partnership including goodwill are treated for the purposes of chargeable gains tax as held by its members as partners. Any dealings by the LLP are treated for those purposes as dealings by its members in partnership and not by the LLP as such. Section 59A(1) makes clear that tax in respect of chargeable gains accruing to the members of the LLP on disposal of any of its assets shall be assessed and charged on them separately. According to an HMRC Statement of Practice, the ownership of the assets is treated by HMRC as being in proportion to the partners’ capital profit shares. Hence Mr Reeves is treated as personally owning 7.4% of the assets and goodwill of BlueCrest.

Although section 2 states that capital gains are only chargeable on UK residents, it is expressed to be subject to section 10. Section 10 provides:

“10 Non-resident with United Kingdom branch or agency

(1)

Subject to any exceptions provided by this Act, a person shall be chargeable to capital gains tax in respect of chargeable gains accruing to him in a year of assessment in which he is not resident and not ordinarily resident in the United Kingdom but is carrying on a trade in the United Kingdom through a branch or agency, and shall be so chargeable on chargeable gains accruing on the disposal—

(a)

of assets situated in the United Kingdom and used in or for the purposes of the trade at or before the time when the capital gain accrued, or

(b)

of assets situated in the United Kingdom and used or held for the purposes of the branch or agency at or before that time, or assets acquired for use by or for the purposes of the branch or agency.

(2)

Subsection (1) above does not apply unless the disposal is made at a time when the person is carrying on the trade in the United Kingdom through a branch or agency.”

It is common ground here that the BlueCrest LLP before it emigrated to Guernsey was a branch or agency within the meaning of section 10. This meant that even though Mr Reeves was a non-resident, he was carrying on a trade in the United Kingdom in respect of assets including goodwill which were situated here. Even after Mr Reeves became non-resident, if BlueCrest had sold its assets he would have been liable to capital gains tax on any chargeable gain arising. At that time, section 10 was the only circumstance in which a non-resident could be liable to a charge on disposal of an asset in the UK.

Importantly, however, the shares held by Mr Reeves in WHR were not assets used in or for the purposes of a trade carried on in the UK and thus were not an asset within section 10.

The proposed emigration of BlueCrest from the UK to Guernsey would trigger the application of section 25 TCGA. Section 25(1) provides that where an asset ceases by virtue of becoming situated outside the United Kingdom to be a chargeable asset in relation to a person, he shall be deemed for all purposes of the TCGA to have disposed of the asset immediately before the time it emigrated and to have immediately reacquired it at its market value at that time. Similarly, section 25(3) provides that where an asset ceases to be a chargeable asset in relation to a person by virtue of his ceasing to carry on a trade in the United Kingdom through a branch or agency he shall be deemed to have disposed of and immediately reacquired the asset at its market value. The effect of section 25 was that if BlueCrest emigrated from being a UK resident LLP to being non-resident at a time when Mr Reeves still held his interest in it then even though he was non-resident he would be deemed to have disposed of and immediately re-acquired his 7.4% interest and would be chargeable to tax on any gain. According to section 25(7), an asset is only a chargeable asset for the purposes of the deemed disposal under section 25(1) or 25(3) if any chargeable gain accruing to the non-resident person on the disposal would be chargeable to capital gains tax under section 10(1).

However, once BlueCrest was owned by WHR, section 10 had no application because WHR was resident in the UK. Since any gain on the disposal would not be chargeable under section 10, section 25 also did not apply and there would be no deemed disposal on the emigration of BlueCrest to Guernsey.

When Mr Reeves gifted his interest in BlueCrest to WHR that was, of course, a disposal of that interest by him to WHR. Even though it was a gift, it is deemed for capital gains tax purposes to be for a consideration equal to the market value of the asset: see section 17 TCGA. The question arising in this appeal is whether Mr Reeves was liable to pay capital gains tax on that disposal of his interest to WHR or whether he can benefit from holdover relief.

The holdover relief provisions are found in sections 165 to 167 TCGA.

“165 Relief for gifts of business assets

(1)

If—

(a)

an individual (“the transferor”) makes a disposal otherwise than under a bargain at arm’s length of an asset within subsection (2) below, and

(b)

a claim for relief under this section is made by the transferor and the person who acquires the asset (“the transferee”) or, where the trustees of a settlement are the transferee, by the transferor alone,

then, subject to subsection (3) and sections 166 and 167, 169, 169B and 169C, subsection (4) below shall apply in relation to the disposal.

(2)

An asset is within this subsection if—

(a)

it is, or is an interest in, an asset used for the purposes of a trade, profession or vocation carried on by—

(i)

the transferor, or

(ii)

his personal company, or

(iii)

a member of a trading group of which the holding company is his personal company, or

(b)

…

(3)

Subsection (4) below does not apply in relation to a disposal if—

(a)

in the case of a disposal of an asset, any gain accruing to the transferor on the disposal is (apart from this section) wholly relieved under Schedule 6, or

(b)

…

(ba) in the case of a disposal of shares or securities, the transferee is a company,

(c)

in the case of a disposal of qualifying corporate bonds, a gain is deemed to accrue by virtue of section 116(10)(b), or

(d)

subsection (3) of section 260 applies in relation to the disposal (or would apply if a claim for relief were duly made under that section).

(4)

Where a claim for relief is made under this section in respect of a disposal—

(a)

the amount of any chargeable gain which, apart from this section, would accrue to the transferor on the disposal, and

(b)

the amount of the consideration for which, apart from this section, the transferee would be regarded for the purposes of capital gains tax as having acquired the asset or, as the case may be, the shares or securities,

shall each be reduced by an amount equal to the held-over gain on the disposal.

(5)

…

(6)

Subject to Part II of Schedule 7 and subsection (7) below, the reference in subsection (4) above to the held-over gain on a disposal is a reference to the chargeable gain which would have accrued on that disposal apart from subsection (4) above and (in appropriate cases) Schedule 6, and in subsection (7) below that chargeable gain is referred to as the unrelieved gain on the disposal.”

It is common ground that Mr Reeves’ interest in BlueCrest was an interest in an asset used by him as transferor for the purposes of trade for the purposes of section 165(2). Further, the interest in BlueCrest, being an LLP, did not come within the carve out in section 165(3)(ba) because it was not a disposal of shares or securities.

The circumstances in which a taxpayer cannot benefit from holdover relief under section 165(4) include those set out in sections 166 and 167.

“166 Gifts to non-residents

(1)

Section 165(4) shall not apply where the transferee is neither resident nor ordinarily resident in the United Kingdom.

(2)

Section 165(4) shall not apply where the transferee is an individual if that individual—

(a)

though resident or ordinarily resident in the United Kingdom, is regarded for the purposes of any double taxation relief arrangements as resident in a territory outside the United Kingdom, and

(b)

by virtue of the arrangements would not be liable in the United Kingdom to tax on a gain arising on a disposal of the asset occurring immediately after its acquisition.

167 Gifts to foreign-controlled companies

(1)

Section 165(4) shall not apply where the transferee is a company which is within subsection (2) below.

(2)

A company is within this subsection if it is controlled by a person who, or by persons each of whom—

(a)

is neither resident nor ordinarily resident in the United Kingdom, and

(b)

is connected with the person making the disposal.

(3)

For the purposes of subsection (2) above, a person who (either alone or with others) controls a company by virtue of holding assets relating to that or any other company and who is resident or ordinarily resident in the United Kingdom shall be regarded as neither resident nor ordinarily resident there if—

(a)

he is regarded for the purposes of any double taxation relief arrangements as resident in a territory outside the United Kingdom, and

(b)

by virtue of the arrangements he would not be liable in the United Kingdom to tax on a gain arising on a disposal of the assets.”

Section 166 did not operate to prevent the application of section 165(4) here because the transferee, WHR, was resident in the United Kingdom and was not an individual.

The definitions of the different concepts used in section 167 are important in this appeal. First, the definition of when the person controlling the company is “connected” with the person making the disposal for the purposes of subsection (2)(b) is defined in section 286 TCGA:

“286 Connected persons: interpretation

(1)

Any question whether a person is connected with another shall for the purposes of this Act be determined in accordance with the following subsections of this section (any provision that one person is connected with another being taken to mean that they are connected with one another).

(2)

A person is connected with an individual if that person is the individual’s husband or wife, or is a relative, or the husband or wife of a relative, of the individual or of the individual’s husband or wife.

(3)

A person, in his capacity as trustee of a settlement, is connected with any individual who in relation to the settlement is a settlor, with any person who is connected with such an individual and with a body corporate which, under section 681 of the Taxes Act, is deemed to be connected with that settlement (“settlement” and “settlor” having for the purposes of this subsection the meanings assigned to them by subsection (4) of the said section 681).

(4)

Except in relation to acquisitions or disposals of partnership assets pursuant to bona fide commercial arrangements, a person is connected with any person with whom he is in partnership, and with the husband or wife or a relative of any individual with whom he is in partnership.

…..

(8)

In this section “relative” means brother, sister, ancestor or lineal descendant.”

The definition of when a company is “controlled” by a person is dealt with in section 288(1) TCGA.

“288 Interpretation

(1)

In this Act, unless the context otherwise requires—

…

“control” shall be construed in accordance with section 416 of the Taxes Act;”

Section 416 is part of Chapter 1 of Part XI of the Income and Corporation Taxes Act 1988 dealing with the taxation of close companies. Chapter 1 contains the interpretative provisions for that Part. A close company is defined in section 414 as one which is under the control of five or fewer participators. Section 416 then defines the meaning of “control”.

“416 Meaning of “associated company” and “control”

(1)

For the purposes of this Part, a company is to be treated as another’s “associated company” at a given time if, at that time or at any other time within one year previously, one of the two has control of the other, or both are under the control of the same person or persons.

(2)

For the purposes of this Part, a person shall be taken to have control of a company if he exercises, or is able to exercise or is entitled to acquire, direct or indirect control over the company’s affairs, and in particular, but without prejudice to the generality of the preceding words, if he possesses or is entitled to acquire—

(a)

the greater part of the share capital or issued share capital of the company or of the voting power in the company; or

(b)

such part of the issued share capital of the company as would, if the whole of the income of the company were in fact distributed among the participators (without regard to any rights which he or any other person has as a loan creditor), entitle him to receive the greater part of the amount so distributed; or

(c)

such rights as would, in the event of the winding-up of the company or in any other circumstances, entitle him to receive the greater part of the assets of the company which would then be available for distribution among the participators.

(3)

Where two or more persons together satisfy any of the conditions of subsection (2) above, they shall be taken to have control of the company.

(4)

For the purposes of subsection (2) above a person shall be treated as entitled to acquire anything which he is entitled to acquire at a future date, or will at a future date be entitled to acquire.

(5)

For the purposes of subsections (2) and (3) above, there shall be attributed to any person any rights or powers of a nominee for him, that is to say, any rights or powers which another person possesses on his behalf or may be required to exercise on his direction or behalf.

(6)

For the purposes of subsections (2) and (3) above, there may also be attributed to any person all the rights and powers of any company of which he has, or he and associates of his have, control or any two or more such companies, or of any associate of his or of any two or more associates of his, including those attributed to a company or associate under subsection (5) above, but not those attributed to an associate under this subsection; and such attributions shall be made under this subsection as will result in the company being treated as under the control of five or fewer participators if it can be so treated.”

The definitions of certain terms used in section 416 ICTA are set out in section 417 ICTA:

“417 Meaning of “participator”, “associate”, “director” and “loan creditor”

(1)

For the purposes of this Part, a “participator” is, in relation to any company, a person having a share or interest in the capital or income of the company, and, without prejudice to the generality of the preceding words, includes—

(a)

any person who possesses, or is entitled to acquire, share capital or voting rights in the company;

(b)

any loan creditor of the company;

(c)

any person who possesses, or is entitled to acquire, a right to receive or participate in distributions of the company (construing “distributions” without regard to section 418) or any amounts payable by the company (in cash or in kind) to loan creditors by way of premium on redemption; and

(d)

any person who is entitled to secure that income or assets (whether present or future) of the company will be applied directly or indirectly for his benefit.

In this subsection references to being entitled to do anything apply where a person is presently entitled to do it at a future date, or will at a future date be entitled to do it.

(2)

The provisions of subsection (1) above are without prejudice to any particular provision of this Part requiring a participator in one company to be treated as being also a participator in another company.

(3)

For the purposes of this Part “associate” means, in relation to a participator—

(a)

any relative or partner of the participator;

(b)

the trustee or trustees of any settlement in relation to which the participator is, or any relative of his (living or dead) is or was, a settlor (“settlement” and “settlor” having here the same meaning as in section 681(4)); and

(c)

where the participator is interested in any shares or obligations of the company which are subject to any trust, or are part of the estate of a deceased person—

(i)

the trustee or trustees of the settlement concerned or, as the case may be, the personal representatives of the deceased; and

(ii)

if the participator is a company, any other company interested in those shares or obligations;

and has a corresponding meaning in relation to a person other than a participator.

(4)

In subsection (3) above “relative” means husband or wife, parent or remoter forebear, child or remoter issue, or brother or sister.”

On 20 December 2011 Mr Reeves and WHR made a joint claim for holdover relief on the disposal by gift of Mr Reeves’ interest in BlueCrest to WHR. That gift had produced a chargeable gain of £33,636,557 because it was treated under section 17 TCGA as having been made for market value. On 26 September 2012 HMRC opened an enquiry into Mr Reeves’ self-assessment tax return for the year ended 5 April 2010. On 2 May 2013 HMRC indicated that they considered the claim to holdover relief invalid because of section 167(2) TCGA. The enquiry was closed on 1 September 2014 and HMRC amended Mr Reeves’ return to disallow the claim. HMRC contended then, as they have contended throughout, that section 167(2) applies to preclude holdover relief because WHR is “controlled” by non-resident persons connected with Mr Reeves, namely Mr Reeves’ wife and children.

Both parties agreed before the FTT and before us that on a literal interpretation of section 167(2) TCGA, HMRC is correct in saying that the disposal does not benefit from holdover relief under section 165(4) because it falls within section 167(2). This is because Mr Reeves’ wife and children are not resident in the United Kingdom. They are clearly connected with Mr Reeves who is the person making the disposal for the purposes of section 167(2)(b). But how is it that they are said to control WHR when they have no interest in it? This results from the operation of section 416(2) ICTA as extended by section 416(6). For this purpose, one reads subsection (2) as follows:

“(2)

For the purposes of section 167(2), Mrs Reeves shall be taken to have control of WHR if she possesses —

(a)

the greater part of the share capital or issued share capital of WHR or of the voting power in WHR.”

But does Mrs Reeves possess the greater part of the share capital of WHR? She does if one attributes to her, pursuant to subsection (6), Mr Reeves’ controlling interest in WHR. One reads subsection (6) as follows:

“(6)

For the purposes of {deciding whether Mrs Reeves possesses the greater part of the share capital of WHR} … , there may also be attributed to Mrs Reeves all the rights and powers of ... any associate of hers, including her husband Mr Reeves but not those attributed to an associate under this subsection; and such attributions shall be made under this subsection as will result in the company being treated as under the control of five or fewer participators if it can be so treated.”

Thus, it is agreed that on its literal interpretation:

Both HMRC and Mr Reeves regard the wording of the relevant statutory provisions as unsatisfactory from their perspective. HMRC contend that the main provision in dispute, section 167(2) TCGA, contains an error in the drafting. They invite us to read in words which would more aptly cover a taxpayer in Mr Reeves’ position whose disposal should not, they submit, benefit from holdover relief. They say that although strictly he is not “connected with the person making the disposal”, section 167(2) should be read as including the situation where the transferee company is controlled by a person who is non-resident and who actually is the person making the disposal.

Unless one can read it like that, HMRC argue, the literal wording creates a loophole from which, fortuitously, Mr Reeves cannot benefit because he has relatives (i) who are associated with him for the purposes of section 416(6), (ii) to whom his rights in WHR can be attributed under that subsection; (iii) who are non-UK resident; and (iv) who are connected with him for the purposes of section 167(2). But the loophole would still exist if Mr Reeves was non-resident himself and had no one connected with him at all or if all his associates were UK resident.

Mr Reeves argues that what is wrong with section 167(2) as expanded by section 416 is that it attributes Mr Reeves’ rights and powers in WHR to his non-resident wife and children even though they have nothing to do with WHR themselves but just happen to be connected to him. This leads to an absurd result because it makes the application of section 167 turn on the happenstance of whether a transferor has an associate - a term which is defined very widely by section 417 - who is non-resident. In particular, it does not matter on this literal interpretation whether Mr Reeves who actually owns the company is non-resident. Even if Mr Reeves and his wife lived in the United Kingdom, he would not be entitled to holdover relief under section 165 if, for example, his parents had retired to Florida or his daughter had taken up a teaching post at an Australian university. His parents or his daughter may be completely unaware of and unconcerned with the existence of this business. They may never have had any dealings with WHR. But the fact that they are associates of his, are connected with him and are non-resident would on the literal interpretation of the provisions mean that his interest would be attributed to them and hence bring the disposal within section 167(2) and outside section 165(4).

Mr Reeves proposes two ways to interpret the legislation to avoid this absurd result. The first way is to rely on the qualification expressed in section 288 TCGA which imports the definition of ‘control’ in section 416 into the TCGA, including into section 167(2). That qualification is that ‘control’ is construed in accordance with section 416 “unless the context otherwise requires”. Mr Reeves submits that the context here does require a modification to the imported definition in order to avoid the absurdity of Mrs Reeves being treated as controlling WHR when she has no connection with it other than being married to Mr Reeves. What the context requires, he submits, is that Mrs Reeves should herself be a participator in WHR within the meaning of section 417(1) before Mr Reeves’ rights and powers are attributed to her under section 416(2).

The second way is to read into section 167(2) the words which are there in section 167(3), namely that for the purposes of subsection (2) a person only controls the company if he or she controls it “by virtue of holding assets relating to that or any other company”. If Mrs Reeves did hold assets relating to WHR, then Mr Reeves accepts that it would be entirely fair for Mr Reeves’ rights and powers to be attributed to her pursuant to section 416(6) so whatever holding she had, however small, would be aggregated with his holding and result in her controlling WHR.

Mr Reeves submits that he can achieve this reading of section 167(2) because the words which are expressly included in section 167(3) must be implicit in the preceding subsection. But if he is wrong in his submission that one can arrive at that result by the ordinary canons of statutory construction, he asserts that there was clearly a mistake by the draftsman which the Tribunal can correct.

Before considering the merits of these different interpretations we consider the case law on statutory construction. In the well-known case of Barclays Mercantile Business Finance Ltd v Mawson [2004] UKHL 51, [2005] 1 AC 684 (‘Mawson’) the House of Lords held that a taxing statute is to be applied by reference to the ordinary principles of statutory construction, i.e. by giving the provision a purposive construction in order to identify its requirements and then deciding whether the actual transaction answers to the statutory description. The question is always whether the relevant provision of the statute, upon its true construction, applies to the facts as found. The Appellant in Mawson, Barclays, claimed a capital allowance for a pipeline which had been purchased and then leased back to the seller. The Inland Revenue’s challenge to the claim for a capital allowance arose from the fact that the transactions formed part of a larger scheme concerning the disposal of the proceeds of sale which Barclays paid for the pipeline. The Special Commissioners and the High Court on appeal held that the expenditure was not incurred for the purpose of Barclays’ finance leasing trade because the transaction lacked commercial reality. That decision was overturned by the Court of Appeal. The House of Lords upheld the Court of Appeal. The statutory requirements for the application of the capital allowance were, they held, concerned entirely with the acts and purposes of the lessor. The relevant provisions said nothing about what the lessee should do with the purchase price, how he found the money to pay the rent or how he used the plant. The findings that the transaction had no commercial reality depended on an examination of what happened to the purchase price after Barclays had paid it. But that did not affect the reality of the expenditure by Barclays.

The role of the court in correcting anomalies created by the literal wording of tax legislation has been considered on many occasions. In Jenks v Dickinson (HM Inspector of Taxes) [1997] STC 853, Neuberger J cited passages from earlier authorities including Mangin v Inland Revenue Commissioner [1971] AC 739 where Lord Donovan had said that the object of the construction of the statute being to ascertain the will of the legislature, it may be presumed that neither injustice nor absurdity was intended. If therefore a literal interpretation would produce such a result, and the language admits of an interpretation which would have avoided it, then such an interpretation may be adopted. Further the history of an enactment and the reasons which led to it being passed may be used as an aid to its construction. Neuberger J went on to refer to two cases where a strained meaning was preferred to the more natural meaning of the words in order to prevent a wholly unreasonable result. He referred in particular to Marshall (HM Inspector of Taxes) v Kerr (1993) 67 TC 56 which dealt with the construction of a deeming provision. The Court of Appeal in that case had said that if the ordinary and natural meaning of the words would lead to injustice or absurdity, “the application of the statutory fiction should be limited to the extent needed to avoid such injustice or absurdity, unless such application would clearly be within the purposes of the fiction”: see the passage from Marshall cited by Neuberger J at page 861 of Jenks. Neuberger J regarded the provision which generated the anomaly in the case before him as also being a deeming provision. He followed the guidance of the Court of Appeal in Marshall that when considering whether one can do some violence to the words and discard their ordinary meaning one can, indeed one should, take into account the fact that one is construing a deeming provision: (page 878)

“This is not to say that ordinary principles of construction somehow cease to apply when one is concerned with interpreting a deeming provision: there is no basis in principle or authority for such a proposition. It is more that, by its very nature, a deeming provision involves artificial assumptions. It will frequently be difficult or unrealistic to expect the legislature to be able satisfactorily to proscribe the precise limit to the circumstances in which, or the extent to which, the artificial assumptions are to be made. … Accordingly, while the rules of construction … apply equally to a deeming provision it is, at least in some circumstances, rather easier to identify a limitation to the ambit of a deeming provision than it is to the ambit of a provision which is not a deeming provision.”

Neuberger J considered that the fact that the anomaly arose from the interrelationship of a deeming provision construed in a literal and unlimited way, an unusual set of facts, and legislation which was complex helped demonstrate why it may be appropriate as a matter of construction to limit the apparently very wide effect of the deeming provision.

Both Mr Reeves and HMRC relied before us on Inco Europe [2000] 1 WLR 586. In that case the House of Lords was considering an application for a stay of High Court proceedings on the grounds that they had been brought in respect of a matter which the parties had agreed to refer to arbitration in the Netherlands. The first instance judge had dismissed the application on the grounds that the arbitration agreement was void. A question arose as to whether the Court of Appeal had jurisdiction to entertain an appeal. Lord Nicholls of Birkenhead recognised that the relevant provision in the Schedule to the Arbitration Act “read literally and in isolation from its context” precluded any right of appeal. His Lordship held that “Several features make it plain beyond a peradventure that on this occasion Homer, in the person of the draftsman … nodded” and that something had gone awry in the drafting. Having regard to the purpose of the provision and its context, that is that it was intended to be a consequential amendment rather than making a major legislative change, he held that the proper interpretation of the provision should give effect to Parliament’s intention. He referred to the court’s role in correcting obvious drafting errors. In suitable cases, in discharging its interpretative function the court will add words, or omit words or substitute words: page 592C-D. However, the power was strictly confined “to plain cases of drafting mistakes”:

“The courts are ever mindful that their constitutional role in this field is interpretative. They must abstain from any course which might have the appearance of judicial legislation. A statute is expressed in language approved and enacted by the legislature. So the courts exercise considerable caution before adding or omitting or substituting words. Before interpreting a statute in this way the court must be abundantly sure of three matters: (1) the intended purpose of the statute or provision in question; (2) that by inadvertence the draftsman and Parliament failed to give effect to that purpose in the provision in question; and (3) the substance of the provision Parliament would have made, although not necessarily the precise words Parliament would have used, had the error in the Bill been noticed. The third of these conditions is of crucial importance. Otherwise any attempt to determine the meaning of the enactment would cross the boundary between construction and legislation …”

Lord Nicholls went on to say that even where these three conditions were met the court may find itself inhibited from interpreting the statutory provision in accordance with what it is satisfied was the underlying intention of Parliament.

Inco Europe was applied by the Court of Appeal in Pollen Estate Trustee Co Ltd v HMRC [2013] EWCA Civ 753, [2013] 1 WLR 3785. In that case the statutory provision appeared to rule out relief from stamp duty land tax for a charity on the purchase price of property acquired by the charity in a situation where it only acquired part of the beneficial interest in the property and not the whole beneficial interest. The literal wording of the charitable exemption meant that although a charity would be entitled to relief from SDLT if it acquired an existing beneficial interest in property as an investment for charitable purposes, it was denied relief if it merely participated in a purchase of a larger interest with non-charities. The Upper Tribunal decision recorded having asked Counsel for HMRC what policy reason there might be for distinguishing the acquisition of an existing undivided share by a charity from the acquisition of part of an undivided share as a result of a joint purchase with a non-charity. One of the reasons given in response was that there was a policy concern that if relief was available where a charity contributed to the purchase price of a property that could lead to avoidance or abuse. Lewison LJ (with whom McFarlane and Laws LJJ agreed) regarded that reason as unconvincing particularly because there was an express provision precluding relief where the transaction had been entered into for the purpose of avoiding the tax: [21]. He therefore held that there was no policy justification for the anomalous position and concluded that it seemed obvious that if Parliament had excluded exemption in such a case, it had done so by mistake. In Pollen Estates the difficulty arose because the exemption applied only where “the purchaser is a charity”. Lewison LJ recognised that simply changing the word “purchaser” to “purchasers” or to “one of the purchasers” did not work. One had to add in the words “to the extent that the purchaser is a charity”. That would have the consequence that a land transaction was partially exempt to the extent of the charity’s interest. He held that that was the correct interpretation even though there was no machinery for determining what part of the interest is held for qualifying charitable purposes; that was not an insuperable objection.

HMRC submit that an unintended loophole exists in section 167(2) because it only applies where the non-resident controller of the transferee company is a person connected with the transferor, not where the controller is the transferor himself. This must be a mistake which should be remedied directly by adding in some words to section 167(2) so that it reads something like:

“(2)

A company is within this subsection if it is controlled by a person who, or by persons each of whom—

(a)

is neither resident nor ordinarily resident in the United Kingdom, and

(b)

[is the person making the disposal or] is connected with the person making the disposal.”

This construction was referred to at the hearing before us as HMRC’s “purposive interpretation”.

The FTT rejected this purposive interpretation. The tribunal accepted Mr Reeves’ argument that section 167 does not address the residence status of the transferor but is solely concerned with the residence of the person to whom the control is transferred. Judge Brooks therefore held that in the circumstances he could not be “abundantly sure” (adopting Lord Nicholls’ phrase from Inco Europe) that Parliament intended that holdover relief under section 165 TCGA should not be available to a non-UK resident transferor making a gift of a business asset to a company he controlled: see [43].

In our judgment, applying the three stage test in Inco Europe, we cannot be abundantly sure what Parliament intended to do about a taxpayer in Mr Reeves’ position, namely one who is a non-resident transferor and who also controls the resident transferee company.

First, one must recognise that the position of such a taxpayer is curious on a number of fronts. Non-resident taxpayers are generally not taxed on their disposals of assets based in the United Kingdom. Section 10 provided at the relevant time for a single exception which applied on the disposal of a narrowly defined class of assets the disposal of which can trigger a chargeable gain even for a non-resident. Section 10 is drafted so that shares in a company do not count as business assets potentially triggering such a charge. This means that even if the company is carrying on a business in the UK, a non-resident shareholder can dispose of those shares without incurring a charge to tax. Thus, when section 10 was enacted it was intended to catch non-resident people who directly owned the business assets but not non-resident people who owned or controlled them via shares in a corporate entity.

Section 10 might also have captured a foreign partnership holding assets. But both parties accepted that the chances of that occurring at the time when the provisions were enacted were very small. Mr Ewart QC appearing for HMRC explained that before 2000, partnerships were either traditional 1890 Act partnerships where everyone had unlimited liability or limited partnerships under the Limited Partnership Act 1907. The former were not popular for those engaged in business and in the latter, the partners who had limited liability could not participate in the business of the partnership; the business was usually conducted by the general partner who was a corporate entity with limited liability.

What has given rise to Mr Reeves’ issue is the superimposition onto this structure of the way in which LLPs were treated in the tax system once they were created in the Limited Liability Partnerships Act 2000. It is the fact that the interest in an LLP is treated, pursuant to section 59A TCGA, as Mr Reeves’ personal asset rather than being allied with an interest in a company that brings his interest in BlueCrest within section 10 and hence means that he needs to rely on the holdover relief in section 165.

As to whether sections 166 and 167 achieve their purpose, that rather depends on how widely one defines the purpose once one tries to ascertain what lay behind the enactment of section 167. Mr Ewart describes the purpose as ensuring that assets which are within the charge to tax are not disposed of in a manner which means that the transferor retains de facto control by gifting the asset to a company which is controlled (within the meaning of section 416(6)) by non-residents so that subsequent disposals of the shares in that company are outside the charge to tax. He said that if section 167(2) cannot be given this purposive construction then sections 10 and 25 TCGA are a completely dead letter. One can circumvent them very simply by doing what Mr Reeves did. He does not even have to set up an offshore trust - he can own the company himself.

Mr Prosser QC, appearing for Mr Reeves, drew our attention to the case of Frankland v IRC [1997] STC 1450 where the Court of Appeal refused to construe a statutory exemption to extend to a transaction which was not covered by its literal wording. Peter Gibson LJ cited the judgment of Oliver LJ in IRC v Sir John Aird’s Settlement Trustees [1983] STC 700, at 707-708 about the dangers of surmising as to the purpose of a statutory provision where a provision construed literally is perfectly intelligible and perfectly capable of operation. Chadwick LJ at page 1464 of Frankland stressed that the purpose of the legislation must be identified in the legislation itself and in any other relevant and admissible material. It is not permissible to speculate as to what the legislature must or might have intended and then strain the statutory language used in order to give effect to that presumed purpose. We were not taken to any admissible material to substantiate HMRC’s submission that the purpose of the provision included catching a non-resident transferor like Mr Reeves.

Mr Prosser describes the purpose of section 167 as to prevent what was known among tax advisers as “the envelope trick”. This was a device by which UK resident taxpayers transferred their assets in the UK to non-resident connected persons in order that subsequent disposals of those shares would not be caught. He says that section 167(2) is designed to prevent the owner of a business asset from doing indirectly what section 166 prevents him from doing directly, namely passing control to a transferee which is controlled by someone other than himself. Section 166 does not contemplate anything other than the asset being transferred to someone else so there is no reason to suppose that section 167 does either.

In our judgment, there is no doubt that the drafting of section 167(2) achieves the purpose for which it was enacted, that is to put an end to the envelope trick as a widely used arrangement by which UK residents could avoid capital gains tax by gifting their assets to a company controlled by non-resident associates. What it does not achieve is sewing up a different, albeit similar, loophole benefiting a non-resident taxpayer who, because of the way in which a combination of other, subsequently enacted, statutory provisions work, would be liable to tax on the disposal of an asset.

If the drafting of section 167(2) had been defective in a way which failed to prevent the envelope trick from working, then that might well have been an error which the Tribunal could repair relying on Inco Europe. But Mr Ewart fairly accepted that Parliament did not address its mind either when section 167 was enacted or when LLPs were created to a non-resident owner of an interest in an LLP like Mr Reeves. We do not consider that the authority of Inco Europe is intended to allow HMRC to extend a provision designed to close one gap so that it closes a different gap which Parliament has not considered.

It may be, as Mr Ewart submits, that other non-resident transferors of assets falling within section 10 would seek to take advantage of the mechanism that Mr Reeves has devised. But if Parliament were then to turn its mind to preventing that, we cannot be abundantly sure how, if at all, it would go about stopping such avoidance and in particular whether it would decide as a matter of policy to treat non-resident transferors who hold section 10 assets because they are partners in an LLP in the same way as it treats non-resident transferors who hold section 10 assets directly.

Mr Prosser pointed out also that the court cannot assume that because the transferor who also holds all the shares in the transferee company is non-resident, that means the effect of the transfer will be to take the gift out of the charge to capital gains tax. He gave the example of the non-resident trustees of a trust which owns a business, part of which they wish to transfer to a UK resident company of which they will own the shares. They may decide to claim holdover relief under section 165(4) pursuant to paragraph 2 of Schedule 7 to the TCGA because although, as non-residents, they would not be liable to tax on any gain themselves, they may wish to prevent the beneficiaries or the settlor from being liable on the gain pursuant to sections 86 and 87 TCGA. There could be no policy reason for refusing holdover relief in that situation since the asset is as much within the charge to tax after as it was before the transaction.

Mr Prosser also relies on the fact that close by sections 165 – 167 is section 162 TCGA. That section provides roll over relief in circumstances where a person who is not a company transfers to a company a business as a going concern together with the whole assets of the business and receives in exchange shares in the company. If the person transferring the business is non-resident and the business constitutes an asset within section 10, the transfer of the business in exchange for shares will enable the non-resident transferor subsequently to sell those shares without paying capital gains tax because the shares in the company do not fall within section 10.

There are obvious differences between the holdover relief available under this provision and the holdover relief available under section 165. Section 162 applies only where all the assets of the business are transferred and section 165 applies only where the transfer is by way of gift. But we accept that section 162 is an instance where Parliament contemplates that a non-resident having an asset within section 10 and transferring it to a company of which he is the shareholder will be entitled to holdover relief even though he will no longer be liable to tax on a gain if he then sells those shares. Of course, the company itself would be liable to tax on the heldover gain if it subsequently sold the business assets. But that is the same under section 165 as under section 162.

Finally, we consider there is some force in Mr Prosser’s submission that it is difficult to believe that an error was made in a short and simply worded provision such as section 167(2) and that the draftsman simply did not notice or appreciate that a reference to a person connected to the transferor would not include the transferor himself.

We therefore reject HMRC’s purposive interpretation as a possible legitimate reading of the provision.

Mr Reeves’ first possible alternative construction is to treat the words that are expressed in section 167(3), namely that the controller has to be someone who controls the company “by virtue of holding assets relating to that or any other company” as being implicit in section 167(2). In the present case if Mrs Reeves and the Reeves’ children had been Treaty non-resident rather than actually non-resident, section 167(3) would not have applied because they do not have control by virtue of holding assets in the company. It is only because they are actually non-resident and the relevant words in section 167(3) are not expressly included in section 167(2) that they appear to be caught and holdover relief is denied.

Mr Prosser submits that section 167(3) cannot be intended to be narrower than section 167(2); it is intended to replicate the effect of section 167(2) for a person who is Treaty non-resident rather than actually non-resident. There can be no rational basis for requiring a link between the associate and the company for the purposes of a Treaty non-resident transferee but not requiring such a link for the purposes of an actually non-resident transferee. Mr Prosser argues that the wording of subsection (3) puts beyond doubt that what Parliament had in mind was control through having an interest in the transferee company.

HMRC argue that the difference in language between the two subsections must be deliberate and that the omission of the reference to control by virtue of having an interest in the company from section 167(2) must mean that section 167(2) control was intended to be wider. Mr Ewart argues that the reason that the words are included in section 167(3) is that the Treaty provisions referred to only operate by reference to particular assets. They could not apply to a person who controls the company without holding any assets.

Mr Prosser drew our attention to O’Rourke (Inspector of Taxes) v Binks [1992] STC 703 as an example of a case where the court transposed words expressly included in some subsections into a subsection from which they had been omitted. In that case the court considered a claim by the taxpayer to make an election under section 72(4) Capital Gains Tax Act 1979 (which was a consolidating Act) in respect of a capital distribution even though it was not a “small” distribution within the meaning of the other subsections of section 72. The problem with the drafting was that although section 72(2) limited its application to small distributions, subsection (4) on its face contained no such limitation and appeared to apply wherever allowable expenditure was less than the amount distributed. The Court of Appeal held that subsection (4) was designed to remedy a potential anomaly created by subsection (2). But if read literally it had an effect that went far beyond merely remedying that anomaly: page 706f. Scott LJ, with whom Stuart-Smith and Lloyd LJJ agreed, said that it was “natural” to read subsection (4) as dealing with only small distributions even though there was no express limitation of its scope. He rejected the strict approach to construction urged by the taxpayer at page 707-8:

“I am unable to accept that the strict approach is the right one. An ambiguity in a statutory provision may arise in more than one way. … The natural construction of sub-s (4) in its context is, in my opinion, that it is limited in its scope. But the absence of any express limitation makes it possible that no limitation was intended. The contrast between the limitation that would be inferred from a reading of the subsection in its statutory context on the one hand, and the absence of any express limitation on the other hand, produces, in my judgment, an ambiguity.

Accordingly, in my opinion, it is permissible as an aid to construction, as an aid to identifying the legislative intention behind s 72(4), to take into account the anomalies that the absence of any limitation to the scope of the subsection will produce. It is permissible to take account of the antecedent legislation. …”

Mr Prosser also referred us to the predecessor provisions to what are now sections 165 onwards of the TCGA. Those were in the Capital Gains Tax Act 1979 into which sections 126A, 126B and 126C were inserted by the Finance Act 1989. They supplemented the existing section 126 which was the equivalent of section 165 giving holdover relief. The first two additional sections, 126A and 126B, then limited that holdover relief in the same way as sections 166 and 167 do. Section 126C was a clawback provision aimed at the emigration of controlling trustees. It provided that where relief had been given under section 126 in respect of a disposal of an asset to a company which was controlled by the trustee of a settlement, and the trustees become non-resident at a time when the company still held the asset and was still controlled by the trustees, a chargeable gain would be deemed to have accrued to the trustees on emigration, that gain being equal to the holdover gain.

Mr Prosser points out that the wording of section 126B was the same as the current section 167. The wording does not specifically state that the trustees control the company by virtue of having an interest in the company rather than simply by virtue of being associated with someone who has an interest in the company. But he says that section 126C is drafted on the assumption that the trustees control the company through some interest in it, however large or small that interest might be. Otherwise it is difficult to see why their emigration to become non-resident should trigger a deemed disposal. The clawback provision only makes sense if the trustees do have an interest in the company and therefore might dispose of that interest after emigrating thereby benefiting from having taken the asset out of charge. Section 126C was not brought forward into the 1992 Act because it was replaced by a general emigration charge for emigrating trustees in what is now section 87 of the TCGA. But Mr Prosser relies on what he says is a presupposition in section 126C that the control by the trustees exists by virtue of them having an interest as supporting his submission that section 126B also presupposes the existence of an interest. Hence the replacement for section 126B, that is now section 167 TCGA, should also be read as having that requirement implicit in it.

We were not convinced by Mr Prosser’s submissions on this point. As we have concluded that the same effect that Mr Reeves sought to achieve by the argument that the words in section 167(3) were implicit in section 167(2) is in fact achieved by the application of the qualifying words in section 288 and by the application of section 3 of the HRA, both discussed below, we do not need to reach a concluded view on this argument.

Section 288 imports the definition of ‘control’ from section 416 ICTA into the TCGA for all purposes, unless the context otherwise requires. The first issue raised by Mr Reeves here is whether the context in which the term ‘control’ is used in section 167(2) does require that it be defined in a different way from the way it is construed in its original place in section 416 ICTA.

There is no doubt that in its original context the term ‘control’ falls to be construed very broadly. This is illustrated by the decision of the House of Lords in R v Inland Revenue Commissioners, ex p Newfields Developments Ltd. [2001] UKHL 27, [2001] 1 WLR 1111 (‘Newfields’). Newfields concerned a claim by the taxpayer company Newfields to small companies’ relief from corporation tax, such relief being subject to a reduction if the company had an associated company. According to section 13(4) ICTA, a company was associated with another company where they were under the control of the same person or persons. ‘Control’ was construed in accordance with section 416. The Revenue Commissioners argued that there were two companies, Newfields and Lawrek, under the control of the trustees of the will trust of W and that both sets of trustees were associates of W’s widow. The rights and powers of both sets of trustees were therefore to be attributed to her so that she was taken to have control of both companies with the result that both were associated.

Lord Hoffmann noted at [10] that although the definition in section 416 starts with a concept of control which reflects its meaning in ordinary speech, “that fairly simple notion is enormously widened by subsequent subsections”. As regards subsection (6), he commented that “[p]lainly the intention of the legislature was to spread the net very wide”.

Mr Prosser appeared in that case for the taxpayer. He argued before the House that the use of the word ‘may’ in the opening words of section 416(6) (in contrast to the word ‘shall’ used in the preceding subsections) meant that the application of the full width of subsection (6) was not mandatory but could be cut back where appropriate. The House of Lords disagreed, holding that the use of the word “may” did not create an elective power but simply allowed for the fact that a wide range of attributions were possible of which only those that would result in a person being treated as in control were required to be made.

The House of Lords also rejected an argument based on the reference in the final words of section 416(6) to ‘five or fewer participators’. Their Lordships held that that part of the provision was relevant only for the original purpose of section 416(6) which was to identify close companies that are subject to a special fiscal regime.

Partial disapplication of section 416

Initially Mr Reeves argued that the whole of subsection (6) of section 416 should be disapplied. However, Mr Prosser accepted at the hearing that HMRC had shown that the subsection has a proper role to play for the purposes of section 167(2). For example, if a non-resident has shares in a holding company and the gift were made by the transferor to a subsidiary, it is right that the non-resident is treated as controlling the subsidiary even though he does not actually have any shares in the subsidiary but only holds shares in the parent company. It is important therefore that those parts of section 416(6) which attribute to a person all the rights and powers “of a company” he controls are imported into section 167(2), so that he is treated as controlling the subsidiary. The modification that Mr Reeves seeks to make to subsection (6) therefore only extends to what Mr Prosser referred to as the “associate attribution rule”, that is those words which attribute to any person all the rights and powers “of any associate” of his.

HMRC argue that the words “unless the context otherwise requires” in section 288 TCGA do not permit the disapplication of part only of section 416. Either the whole of section 416 including subsection (6) applies unmodified to define ‘control’ as it is used in a provision in the TCGA or it does not, in which case there must be some other definition of ‘control’ adopted by the TCGA. There is no definition of ‘control’ provided by Parliament for section 167 other than the cross reference in section 288 to section 416. The disapplication of the whole of section 416 would therefore leave the term as used in section 167 without definition.

We do not accept that the qualification in section 288 to the importing of section 416 has to be on an all or nothing basis as HMRC submit. Mr Prosser referred us to Bennion on Statutory Interpretation, 6^th ed., p 523 which discusses the use of the phrase “unless the contrary intention appears” and states that “A contrary intention may apply to a part only of the definition”. The example by given Bennion is the case of Starke and another (executors of Brown deceased) v Inland Revenue Commissioners [1995] STC 689 (‘Starke’). In that case the question arose whether the Interpretation Act 1978 required the word ‘land’ used in the Inheritance Tax Act 1984 provisions which defined agricultural property as including ‘agricultural land or pasture’ to be construed as including ‘buildings or other structures’. The executors of the estate relied on the Interpretation Act to argue that the reference in the exemption to ‘land’ did include buildings and other structures. This meant that the farmhouse included in the estate benefited from agricultural property relief. HMRC argued that a contrary intention was apparent from the rest of the definition of ‘agricultural property’ which expressly included certain cottages, farm buildings and farmhouses provided they had particular features.

Morritt LJ in Starke noted that it was plain that the definition from section 5 of the Interpretation Act was not excluded in its entirety. He noted also that the form of the definition in the Interpretation Act provided that the word ‘land’ includes specified items, not that the word means such items. He said: “In such a case I can see no reason why the contrary intention referred to in s 5 may not appear in respect of some only of the specified items but not others”. The real question was thus whether the definition of agricultural property showed an intention that the word ‘land’ used there should not include ‘buildings and other structures’. Given that the Inheritance Tax Act definition of ‘land’ specifically referred to certain kinds of buildings, he held that the judge had been right to conclude that that element of the Interpretation Act definition of ‘land’ did not apply.

We recognise that Starke is not on all fours with the present case. This is not a case where the form of definition used provides that ‘control’ includes particular situations but rather that it has a meaning which is extended by the mandatory attributions in the subsections. Further, there was in Starke clear wording in the rest of the Inheritance Tax Act provision that made it nonsensical to include buildings in the definition of land in that context.

However, here a statute imports a complex, many-limbed definition from another statute and provides that the definition applies in a number of different places in the importing statute. We do not see that it makes sense to treat the qualifying words that Parliament has included in section 288 as requiring a choice between applying every element in the definition or none. We derive some support for this conclusion from the speech of Lord Hoffmann in Newfields to which we have already referred. Lord Hoffmann disagreed with the decision of the Court of Appeal which had held, relying on the closing words of section 416(6), that unless the attributions of rights and powers would result in the company being treated as under the control of five or fewer participators, the attributions could not be made at all: [26]. Lord Hoffmann said:

“27 This construction produces a very arbitrary result and would appear to make section 416(6) an unsuitable element in any definition of control for the purposes of section 13(4). …. In deciding what counts as controlling another company, it would be illogical to attribute additional powers only if the effect was to bring that other company within the definition of a close company. That would seem an irrelevant consideration. Mr Prosser said that one could avoid this illogicality by treating the whole of subsection (6) as applicable only to the question of whether a company was a close company. But this would leave it open to anyone to claim small companies’ relief by dividing his business between companies controlled by himself and his wife … . The absence of the attribution provisions of subsection (6) would leave a large gap in the defence which section 13(4) provides for the public revenue.”

He therefore held that the closing words of subsection (6) should not be treated as part of the definition of ‘control’ when considering how that definition is to apply in a context where it is used for a purpose other than the purpose for which it was enacted. The qualification provided by those closing words had “no relevance to any case in which the general definition of control, as set out in the rest of section 416(2) to (6) is sufficient answer to the statutory question”.

Newfields therefore recognises that where a definition has been drafted for one purpose but is imported into another context for a different purpose one must take care in applying it in that other context not to include unsuitable elements that arrive at results which are illogical or which require one to take into account an irrelevant consideration. Again, we recognise that this case is not on all fours with Newfields because there is no doubt that the associate attribution rule is part of the definition of ‘control’ and not some additional wording which is intended only to apply in the original context of defining close companies. But it would not be possible to follow Lord Hoffmann’s guidance if one were faced with the choice of either applying section 416 in its unmodified entirety or disapplying it.

We also consider the present case has parallels with the provisions considered by Neuberger J in Jenks referred to earlier. Section 416(6) deems a person to be a controller of a company in certain circumstances. That brings into play the principles set out in Marshall that it is important to keep the artificiality created by a deeming provision within the bounds which Parliament must have intended. Again, we consider that this supports an interpretation of the section 288 caveat to the import of section 416(6) into section 167(2) which can be more nuanced than HMRC submit.

Does the context “otherwise require”?

We have already described the anomaly that is created by the literal wording of section 416(6), namely that it attributes to a non-resident connected person who has no interest in the transferee company the rights and powers of the owner of the transferee, regardless of whether the transferor is resident or non-resident. It makes the availability of relief dependent on a factor which must be irrelevant in any sensible consideration of whether relief ought or ought not to be available.

HMRC argue however that the associate attribution rule serves two useful purposes. The first is based on specific cases where the rule is needed to prevent arrangements which should be caught from benefiting from holdover relief. The second was based on the need for a bright line to be drawn in a provision aimed at preventing tax avoidance and for that line to be drawn in such a way that a wide range of situations will be on the side of the line that results in holdover relief being denied.

The example HMRC gave of an arrangement which needs to be caught by section 167(2) is given by the FTT at [18], namely where a company is owned by someone who is UK resident but whose wife is non-UK resident. The UK resident could transfer his business assets – say an asset of the kind that would fall within section 10(1)(a) or (b) - to the company claiming holdover relief under section 165 and then give his interest in the company to his non-resident wife. That transfer would not be subject to tax because of the general automatic holdover in section 58 TCGA applying to transfers from one spouse or civil partner to the other. The non-resident spouse could then sell the interest in the company free of tax. If the UK resident had just given the assets to his non-resident wife, then any later disposal by her would be caught by section 10 and within the charge to tax. But if he first gives the assets to a UK company and then transfers the interest in the company to his non-resident wife, the subsequent transfer would not be caught (since shares in a company are not assets within section 10).

We agree with Mr Prosser that this rather far-fetched example does not provide any justification for the breadth of the associate attribution rule. Section 58 applies to transfers between spouses regardless of their residence status. As Mr Prosser pointed out, in 1988, Parliament amended the spousal transfer relief so that it applies where one spouse is UK resident for tax purposes and the other is not, but they are not permanently separated. Parliament was prepared as a matter of policy to tolerate assets being taken out of charge by transfer under section 58, presumably on the basis that it will be rare that spouses who have not separated are nevertheless tax resident in different jurisdictions. If that is the case, Mr Prosser asks, why should Parliament be concerned about the supposed device contemplated in the example given by HMRC? We also accept his submission that it is unlikely that Parliament intended the associate attribution rule to prevent holdover relief being claimed under section 165 by a UK resident husband transferring his asset to a UK resident company simply because he might later transfer his interest to his non-resident wife. He might have no intention of ever transferring the interest to her, so it would be unfair to deprive him of the holdover relief just in case he later did so. If Parliament was really concerned about this possibility, it is more likely to have dealt with it by enacting some clawback provision if and when the husband does try to make the transfer.

Mr Prosser described the associate attribution rule as a sledgehammer to crack a nut, that nut being the supposed device suggested by HMRC. Further, the associate attribution rule is not limited to where the associate is the wife of the transferor. The example does not provide any justification at all for the inclusion of the parents and children or remoter forebears and issue of the transferor in section 416(6). We agree with Mr Prosser that HMRC have been unable to suggest a plausible situation in which the associate attribution rule would be a useful provision in preventing tax avoidance.

Mr Ewart then seeks to justify the breadth of section 167(2) read in conjunction with section 416(6) by the need for tax avoidance measures to be drafted widely and to create a bright line between those who are caught and those who are not caught. He submitted that it is impossible for HMRC to predict how ingenious tax advisers will devise schemes to exploit advantages if they can. He set out for us the history of what he described as the battle between HMRC and Parliament on the one side and taxpayers and their advisers on the other, in particular in their use of companies with non-resident shareholders or off-shore trusts. That is why Parliament must provide for as wide as possible an application to cover all possible schemes.

This was the argument that found favour with the FTT. Judge Brooks recognised that it appeared somewhat irrational and illogical to preclude holdover relief because of the attribution of ‘control’ in a company to a non-UK resident child who has absolutely no interest in the company or its operation. However, he referred to the dictum of Morritt LJ in Steele (Inspector of Taxes) v EVC International NV (formerly European Vinyls Corp (Holdings) BV) [1996] STC 785, to the effect that anti-avoidance provisions do sometimes have a greater scope than strictly required with possible unforeseen and unwelcome consequences. He was thus unable to conclude that Parliament necessarily intended control for the purposes of section 167(2) TCGA to refer to ‘real’ as opposed to ‘fictional’ control: [37] and [38].

As regards the alleged absurdity of the result of the literal interpretation, Mr Ewart argued that the attribution of Mr Reeves’ rights and powers to Mrs Reeves would be no less ‘absurd’ or ‘anomalous’ if she owned a 1% interest in WHR. Whether she owned 1% or no interest in WHR, she could not remotely be described as controlling WHR in the sense of being able to determine its decisions or influence its policy. He contrasted the way in which the definition of ‘control’ is drafted in section 416 with a very different kind of provision, for example in section 840 ICTA. That adopts a definition of control which depends on the affairs of the company being conducted in accordance with the wishes of the putative controller – a definition closer to the meaning of control in ordinary speech.

We fully recognise the legitimacy of drafting techniques that set bright lines to achieve clarity and minimise the need for discretion or value judgements in enforcement. We do not therefore consider that a provision which removes holdover relief where the non-resident spouse owns some interest in the company however small can be regarded as arbitrary for this reason. Such a bright line avoids the difficulty of fixing some random shareholding limit of say 5%, leaving taxpayers whose non-resident associates own 6% to complain that they are really in no different position so far as control of the companies concerned. We accept that section 416 adopts a different drafting approach from the approach in, for example, section 840 ICTA and that Parliament’s choice here must be respected.

But we do not accept the argument that to rely on a non-resident associate who has a very small interest in the transferee is just as absurd or unfair as relying on a non-resident associate who has no interest at all. It is true that the ownership of a few shares in the transferee may not result in any real control, but ownership of some interest however small is a commonly used “bright line” in tax affairs. The holding of assets by a non-resident associate in the transferee company is a fact which can be ascertained in straightforward ways. It is likely to be the result of a deliberate decision on the part of the transferor to structure the transferee company in a particular way so that both the transferor and the associate are aware of it and can take it into account in their legitimate tax planning. The ownership of the interest, if it threatens to jeopardise the availability of holdover relief can be easily reversed by requiring that associate to dispose of his interest in the transferee before the transaction takes place so that section 167(2) is not applicable.

We are satisfied that it cannot have been Parliament’s intention that holdover relief should be withheld merely because there is someone who is connected to the transferor and is non-resident even if they have no interest in the transferee company. We therefore hold that the context in which the word ‘control’ is used in section 167(2) does require some modification of section 416, pursuant to section 288, when the definition is imported into the TCGA for the purposes of section 167(2).

What modification is required by the context in which ‘control’ is used in section 167(2)?

Since we are satisfied that the context of section 167(2) into which the definition of control in section 416(6) is imported does require the modification of section 416(6), what should that modification be? During the course of argument, Mr Prosser accepted not only that it would be going too far to disapply the whole of subsection (6) but also that the associate attribution rule should apply where the non-resident associate has an interest in the transferee company.

An illustration of how this was done by the legislature can be found in section 96(10) TCGA. That reads: (emphasis added)

“(10)

For the purposes of this section –

(a)

the question whether a company is controlled by a person or persons shall be construed in accordance with section 416 of the Taxes Act, but in deciding that question for those purposes no rights or powers of (or attributed to) an associate or associates of a person shall be attributed to him under section 416(6) if he is not a participator in the company;

(b)

“participator” has the meaning given by section 417(1) of the Taxes Act.”

Mr Prosser described the legislative history of this provision. A draft of the clause without those words in bold was introduced by the Finance Bill 1991 to serve as the definition of control for the purposes of taxing the beneficiary of an offshore trust who receives a benefit in circumstances where the trustees have made or subsequently make a gain. The Bill proposed to extend that to a case where the trustees gave the benefit not to the beneficiary directly but to a company controlled by the beneficiary. The proposed clause provided that the beneficiary was to be treated as having received that benefit if it was received by a company he controlled. The proposed clause was initially drafted as stating simply that the question whether a company is controlled by a person or persons should be construed in accordance with section 416.

When the Bill was going through Standing Committee, the responsible Government Minister, Francis Maude, referred to a proposed amendment to the definition of ‘control’. He described the definition of ‘control’ as providing that not only can a person’s own rights and powers be taken into account in determining whether he or she controls the company but also those of any associates. That term includes a person’s relatives and the trustees of the settlement of which he or she is the settlor or beneficiary. Mr Maude said that the Government had received representations to the effect that this could result in payments by trustees being attributed for tax purposes to individuals who have no interest in the company. That would, he said, go wider than the Government intended. The clause was therefore amended to limit the definition of ‘control’ so that only those who have an interest in the company as a participator will fall within the provisions.

The difficulty with transposing that solution into the present situation is that the import of section 416 is not achieved by a statutory provision specifically directed at the meaning of ‘control’ for the purposes of section 167 but is a compendious importing provision applying to all uses of the word ‘control’ in the TCGA. In the present statutory framework, it is certainly not appropriate to add in those qualifying words to section 288 so that they qualify the import of section 416 in every instance where the word ‘control’ is used in the TCGA. The qualification is needed, for the purposes of this appeal at least, only insofar as that definition is imported by section 288 for the purposes of section 167(2).

Given the different statutory structure this is most easily done by construing section 167(2) as including the words that are included in section 167(3). It seems to have been common ground between the parties in this case that the effect of that is identical to the effect of the qualification added into the clause in the Finance Bill 1991 which ultimately became section 96(10) TCGA.

We therefore hold that the context of section 167(2) requires that pursuant to section 288, the artificial assumptions to be made (to use Neuberger J’s phrase in Jenks) by the attributions of interests between associates within section 416 are limited to connected persons who control the transferee by virtue of holding assets relating to that or any other company. It follows that Mr Reeves is not precluded from benefiting from holdover relief because his wife and children, who have no interest in WHR, are non-resident and connected with him.

Mr Reeves relies on his rights under the ECHR in the event that we decide that the literal interpretation of section 167(2) prevails and therefore that holdover relief is denied if there is a non-resident person connected to the transferor, even if that person has no shareholding or other interest in the company. Mr Prosser accepted that ECHR rights would not assist Mr Reeves if the Tribunal had accepted HMRC’s purposive interpretation of section 167(2) stretching it to apply where the transferee company is controlled by the non-resident transferor as well as where it is controlled by non-resident persons connected to him.

Section 3 of the Human Rights Act 1998 provides that so far as it is possible to do so primary legislation, whenever enacted, must be read and given effect in a way which is compatible with the Convention rights. Section 1 of the HRA defines the Convention rights as including A1P1 and Article 14 ECHR.

Article 14 ECHR provides:

“Prohibition of discrimination

The enjoyment of the rights and freedoms set forth in this Convention shall be secured without discrimination on any ground such as sex, race, colour, language, religion, political or other opinion, national or social origin, association with a national minority, property, birth or other status.”

A1P1 provides:

“Protection of property

Every natural or legal person is entitled to the peaceful enjoyment of his possessions. No one shall be deprived of his possessions except in the public interest and subject to the conditions provided for by law and by the general principles of international law.

The preceding provisions shall not, however, in any way impair the right of a State to enforce such laws as it deems necessary to control the use of property in accordance with the general interest or to secure the payment of taxes or other contributions or penalties.”

It was common ground that A1P1 is engaged by Mr Reeves’ right not to be taxed by being deprived of holdover relief by section 167. As Mr Ewart expressed it, HMRC accepts that there has been a deprivation of Mr Reeves’ property here, namely a deprivation of the money which Mr Reeves would have to pay by way of tax on the disposal of his interest in BlueCrest to WHR if he and WHR cannot claim holdover relief under section 165.

A1P1 expressly refers to the right of a State to secure the payment of taxes. But it was accepted by both parties that an interference with possessions must strike a fair balance between the general interests of the community on the one hand and the requirement of the protection of an individual’s rights on the other. The test to be applied by the court when considering Article 14 in conjunction with A1P1 was recently considered by the Supreme Court in In Re Brewster [2017] UKSC 8, [2017] 1 WLR 519. The Court was considering a challenge to the provisions of a public service employees’ pension scheme which made the availability of a survivor’s pension to an unmarried partner of the deceased scheme member dependent on the deceased having previously nominated the unmarried partner. No such requirement for nomination was imposed in respect of married survivors. Lord Kerr of Tonaghmore JSC, giving the judgment of the Court, set out the following principles:

Lord Kerr was prepared to accept for the purposes of that appeal that the Court should respect the legislature’s policy choice unless it was “manifestly without reasonable foundation”. A future case may, he said, determine that the test requires adjustment to cater for the situation where the proffered reasons are ex post facto. But that did not affect the outcome of the appeal because the Court was satisfied that the discriminatory provision was indeed manifestly without reasonable foundation.

The first issue to be addressed in this appeal is whether the provisions of the TCGA are discriminatory because they treat Mr Reeves differently on the basis of the fact that his wife and children are non-resident from how they would treat Mr Reeves if his wife and children were resident in the UK. It is common ground that place of residence constitutes an aspect of personal status for the purposes of Article 14: see e.g. Carson and others v United Kingdom appn 42184/05 at [61] – [62] and the cases cited there. Mr Reeves relies on his status as a person married to a non-resident or being the father of non-residents.

HMRC argued that this was an impermissible extension of the status requirement because Mr Reeves was not relying on some characteristic of himself but rather on his association with someone whose status was relevant. However, it is clear from the judgment of the ECtHR in Guberina v Croatia (2018) 66 EHRR 11 that this objection must be rejected. In that case the applicant complained of discrimination based on the disability of his child. The applicant had sought a tax exemption for the purchase of a house. The relevant legislation provided that the exemption applied where a person bought a house to resolve his housing needs. The exemption was refused on the basis the applicant himself had no particular special needs. In [67] onwards, the ECtHR set out the general principles applicable:

Turning to the case before it, the ECtHR said (omitting citation of authorities):

“77.

The present case concerns a situation in which the applicant did not allege discriminatory treatment related to his own disability but rather his alleged unfavourable treatment on the basis of the disability of his child, with whom he lives and for whom he provides care. In other words, in the present case the question arises to what extent the applicant, who does not himself belong to a disadvantaged group, nevertheless suffers less favourable treatment on grounds relating to the disability of his child. …

78.

In this connection the Court reiterates that the words “other status” have generally been given a wide meaning in its case law (…) and their interpretation has not been limited to characteristics which are personal in the sense that they are innate or inherent. (…) Accordingly, for instance, a question of discrimination arose in cases where the status of the applicants, which served as the alleged basis for discriminatory treatment, was determined in relation to their family situation, such as the place of residence of their children (…). It thus follows, in the light of its objective and nature of the rights which it seeks to safeguard, that Article 14 of the Convention also covers instances in which an individual is treated less favourably on the basis of another person’s status or protected characteristics.”

In the present case, there clearly is discrimination on the basis of Mr Reeves’ status as a person with a non-resident wife and children as compared to a person with a resident wife and children. If Mr Reeves’ wife and children had been resident in the UK then the attribution to them of his controlling interest in WHR would have made no difference to his ability to benefit from holdover relief under section 165. It is only because they are non-resident that the attribution to them of his controlling interest would preclude such reliance on the literal interpretation of section 167(2) read together with section 416(6).

Mr Ewart argued that although it is true that the application of the literal interpretation of section 416(6) has this effect, the result in this particular case is not irrational because of Mr Reeves’ own non-resident status. In other words, HMRC argue that Mr Reeves in some general sense ought not to benefit from holdover relief because he is non-resident, so that the literal interpretation as it applies in his case is not devoid of rational foundation. We do not regard that as a legitimate point to raise. Mr Reeves like every other person whether resident or non-resident is entitled to insist on being taxed only in accordance with the law, including in accordance with his rights under the Convention. It is accepted that Mr Reeves is properly a “victim” of the legislation and that the legislation itself does not treat as relevant the status of the transferor; it is only concerned with the residence of those controlling the transferee. One must look at the effect of the legislation more generally not simply in the context of the particular facts of the claimant: see Springett v United Kingdom (Appns 34726/04, 14287/05, 34702/05, decn of 27 April 2010) at page 5. The fact that the claimant is a non-resident transferor cannot affect the validity of the rule under Article 14 or A1P1 when the rule applies equally to resident and non-resident transferors.

Similarly, we do not accept Mr Ewart’s submission that the relevant provision to consider here is the taxing provision in section 2 TCGA. The tribunal cannot look at section 2 in isolation. In any event, the wording of section 2 does not assist HMRC because it is itself expressed as being “subject to any exceptions provided by this Act”. Further, section 2 only charges capital gains tax in respect of gains accruing in a year during any part of which Mr Reeves is resident in the United Kingdom. It is not that section which imposes the tax charge on Mr Reeves but rather section 10 which is also expressed to be “subject to any exceptions provided by this Act”. Thus it is clear that in determining whether Mr Reeves has been the victim of discrimination it is essential to look at all the statutory building blocks which lead to the imposition of this particular charge. It is the combination of provisions which operates to deprive him of his possession.

The First-tier Tribunal dealt with the human rights claim at paragraph 72 of its decision. The FTT gave brief reasons for rejecting the human rights claim namely that “Mr Reeves has not been treated differently to any other person with a non-UK resident wife and children and therefore has not been subject to any discrimination.” That reasoning cannot stand. Mr Reeves is not arguing that he has been treated differently from all those who are in precisely the same position as him but rather that he is one of that class and that class has been treated differently from people with UK resident spouses and children. To say that there are other people in the same class, who are also discriminated against, is no basis for concluding that no discrimination has taken place.

We turn therefore to the question of whether HMRC has shown that the discriminatory treatment of Mr Reeves is justified and proportionate. The test for the claimed justification for a difference in treatment was said by Lord Kerr in In Re Brewster to be well settled: [66]. He referred to the test laid down by Lord Reed JSC in Bank Mellat v HM Treasury (No 2) [2013] UKSC 39, [2014] AC 700 as requiring the court to determine (1) whether the objective of the measure is sufficiently important to justify the limitation of a protected right, (2) whether the measure is rationally connected to the objective, (3) whether a less intrusive measure could have been used without unacceptably compromising the achievement of the objective, and (4) whether, balancing the severity of the measure’s effects on the rights of the person to whom it applies against the importance of the objective, to the extent that the measure will contribute to its achievement, the former outweighs the latter.

Here Mr Reeves accepted that the application of section 167(2) and section 416 is justified and proportionate insofar as it treats the rights of associates as attributed to each other where the associates have some interest in the transferee company however small. But he argues that the measure, if interpreted literally, is too intrusive and is not connected to the achievement of the objective of countering tax avoidance.

It was accepted by Mr Ewart that this case was similar to Brewster because no thought had been given by the decision-maker to the appropriateness of including an associate who has no interest in the transferee company in the class of connected people for the purposes of section 167(2). It was not part of HMRC’s case that Parliament either in 1989 or in 1992 in enacting these provisions was thinking expressly about the section 10 situation.

Many of the same arguments put forward in relation to Mr Reeves’ reliance on the qualifying words of section 288 were also prayed in aid on the question of the proportionality and justification of the provisions for the purposes of Article 14.

We accept that although tax avoidance in a broad sense is a legitimate aim of the provision, the literal interpretation of section 167(2) fails on proportionality grounds because of the anomalous position it creates. The importance of bright lines as a justification for discriminatory treatment in the context of Article 14 was also considered by the Supreme Court in Brewster. Lord Kerr cited at [60] a passage from the joint dissenting judgment of Lord Sumption and Lord Reed JJSC in R (Tigere) v Secretary of State for Business, Innovation and Skills (Just For Kids Law intervening) [2015] 1 WLR 3820. They described the advantages of a clear rule as being capable of being applied accurately and consistently, simplifying administration to enable speedy decisions to be made particularly where there is a need to process a very large number of applications within a short time. Lord Kerr contrasted the position in Tigere with the position in the instant case where no thought was given to possible difficulties with administration that might arise if the surviving cohabitee nomination procedure was not included in the new scheme. Further the respondent had not been able to produce tangible evidence that there would be significant problems if the requirement was abandoned: “Vague suggestions as to the workability of the scheme and the advantages of actuarial predictions were made but these were not supported by evidence”: [62].

In our judgment, it is impossible to justify the application of section 167(2) by reference to the residence or non-residence of a spouse who holds no interest in the company to which the gift is being transferred. Indeed, the application of the provision in circumstances where the taxpayer happens to have a relative living abroad strikes us as precisely the kind of provision at which Article 14 in conjunction with A1P1 is aimed. First, its application is entirely unexpected. No UK resident taxpayer transferring a gift to a company which is wholly owned by other UK resident shareholders would expect to have any difficulty with relying on section 165(4) because of the operation of section 167(2). They are unlikely to seek advice in relation to the issue. It would be a very sharp-eyed adviser who would think of checking with his client the tax residence status of all his parents and grandparents, children and grandchildren, brothers and sisters. It operates as a potential trap for taxpayers and their advisers when trying to plan for the future of their business.

Secondly, its application is entirely arbitrary because there is no reason why that taxpayer should be treated differently depending on the answers to those questions.

Thirdly it may well be the case that many taxpayers over the years have claimed holdover relief in their self-assessment, this has not been queried by HMRC and they have successfully relied on section 165 when in fact on the literal interpretation they were not entitled to do so. It is therefore a provision which in fact places a discretion in the hands of a taxing authority to choose in respect of which taxpayers’ self-assessments it will open an enquiry in the expectation that in many cases there will be a non-resident associate within the taxpayer’s wider family whom they can point to, thereby depriving that particular taxpayer of holdover relief. We are not of course suggesting that HMRC have, in this instance, behaved in such a manner or that they ever would. But this is a provision which is open to abuse by a taxing authority which might be so minded. The taxing authority can let most self-assessments go by without question but pick out one taxpayer and insist that he provide them with wide ranging information about the residence and tax affairs of all the members of his family or risk forfeiting what is often a very valuable relief. It would be impossible for the taxpayer to plan his affairs to benefit from the relief in many cases where Parliament clearly intended that he should benefit. Instead of simply having to prevail on his relative to sell their stake in the transferee company he would have to try to persuade his parents to come back from Florida or his daughter to leave her post at the Australian university in order to benefit from holdover relief.

Mr Ewart submitted that this is a general problem with the tax system which relies on self-assessment by taxpayers. The same problem would arise with someone who claims holdover relief in his self-assessment ignoring the fact that a relative living abroad has a few shares or other small interest in the company. But that takes us back to the justification of a bright line between those associates who have an interest however small in the transferee company and those who have no such interest. Everyone will know that the relative has an interest in the company and would not be surprised if the residence of that person is relevant to a tax matter.

We therefore hold that if the literal interpretation of section 167(2) read together with section 416(6) prevails, it infringes Mr Reeves’ rights under Article 14 of the European Convention in conjunction with A1P1. The legislation operates to deprive him of his possessions (namely the tax that he would have to pay because he cannot rely on holdover relief in section 165) on a discriminatory basis, namely because his wife and children are non-resident in the UK rather than resident in the UK. It is unjustified and disproportionate in circumstances where his wife and children have no interest in the asset the transfer of which has triggered the charge to tax.

In the light of that finding we do not need to consider whether the legislation also infringes A1P1 on its own rather than in conjunction with Article 14.

We then move to whether section 167(2) can be “read down” pursuant to section 3 of the Human Rights Act 1998. We have reminded ourselves of the test laid down by the House of Lords in Ghaidan v Godin-Mendoza [2004] UKHL 30. Lord Nicholls of Birkenhead in Ghaidan reiterated that the operation of section 3 does not depend on there being an ambiguity in the legislation in question: even if there is no doubt as to its meaning according to the ordinary principles of interpretation, section 3 may nonetheless require it to be given a different meaning: see particularly [32] and [33].

In our judgment, it is possible to read section 167(2) in a manner which is compliant with Mr Reeves’ rights and is not inconsistent with a fundamental feature of the legislation. One can do this by reading into section 167(2) the wording in section 167(3). In other words, if we are wrong that it is possible to construe section 167(2) as requiring the connected person to have an interest in the transferee company using the ordinary domestic canons of construction (see [95] and [96] above) then one can certainly do so using the more extensive powers conferred by section 3 HRA.

Mr Reeves also argued that section 167(2) whether construed purposively as HMRC contended or construed literally was a breach of the TFEU provisions on the free movement of capital. This raised five sub-issues:

In the light of our decision that Mr Reeves succeeds on the proper construction of section 167(2) and on the application of Article 14 ECHR in conjunction with A1P1 we do not have to address these issues. We had the benefit of clear and very helpful submissions on these points by Mr Yates on behalf of Mr Reeves and by Ms Abram on behalf of HMRC. We mean no disrespect to either of them in declining to arrive at a decision on these issues which are complicated and on which we consider the answers to be not at all clear. We do not necessarily endorse the conclusions arrived at by the FTT, but we do not consider it appropriate to embark on the legal analysis which would be necessary to determine those five issues.

The appeal is therefore allowed.

Any application for costs in relation to this appeal must be made in writing within one month after the date of release of this decision and be accompanied by a schedule of costs claimed with the application as required by rule 10(5)(b) of the Tribunal Procedure (Upper Tribunal) Rules 2008.