HM Revenue & Customs v Prizedome Ltd & Anor

This is an appeal by The Commissioners for Her Majesty's Revenue & Customs (“HMRC”) against a decision (“the Decision”) released on 16 May 2007 of the Special Commissioners allowing appeals by Limitgood Ltd (“Limitgood”) and Prizedome Ltd (“Prizedome”) (together “the respondents”) against amendments to their corporation tax self-assessments for the periods ending 30 September 2001 and 30 September 2002 and by Limitgood only for its period ending 31 December 2004. The appeals were allowed by the casting vote of the chairman, Mr Theodore Wallace. His fellow commissioner, Dr John Avery Jones, would have dismissed the respondents’ appeals.

The appeals concern the proper construction of provisions contained in schedule 7A to the Taxation of Chargeable Gains Act 1992 (“the 1992 Act”). Schedule 7A was inserted by section 88 of the Finance Act 1993 and schedule 8 to that Act. Broadly stated, the aim of schedule 7A was to restrict the ability of a company that joined a group to utilise unrelieved losses (“pre-entry losses”) which it had already realised or which it subsequently realised on assets that it owned when it joined the group in question by specifying the gains from which such losses could be deducted.

The essential facts, which were agreed and are set out in the Decision, are as follows. (Except as otherwise stated in this judgment references to statutory provisions are to provisions in the 1992 Act.) What follows is taken very largely from paragraphs 4 to 11 of the Decision.

On 26 September 2000 Limitgood and Prizedome each acquired one-half of the issued share capital of Coalite Group Ltd (“Coalite”) from Anglo United plc (“Anglo”) for a consideration of £1 each. As at that date Coalite was worthless. Both Limitgood and Prizedome were wholly-owned subsidiaries of Anglo. The capital gains tax base cost for Anglo of the Coalite shares was £486,949,498. Under section 171(1) the deemed acquisition costs for Limitgood and Prizedome was £243,474,749 each.

On 27 September 2000, a wholly-owned subsidiary of Grantchester Holdings plc (which changed its name to Grantchester Ltd on 19 December 2000 and is referred to as “GL”) acquired the shares in Limitgood and Prizedome for a combined consideration of £4 million. On ceasing to be members of the Anglo Group, Limitgood and Prizedome were each deemed to have disposed of their holdings in Coalite at market value under section 179(3) and thus to have realised capital tax losses under section 179(4). Those losses were pre-entry losses of Limitgood and Prizedome in relation to the GL Group within paragraph 1(2) of schedule 7A. Limitgood and Prizedome each claimed losses of £113,921,249 in their corporation tax self-assessment returns for the accounting period to 30 September 2000, being limited by reason of depreciatory dividends and sections 176 and 177.

On 12 October 2000 other companies in the GL chargeable gains group realised gains totalling £28,956,478 which were treated as accruing to Limitgood by reason of elections under section 171A.

On 19 December 2000 the issued share capital of GL was acquired by Grantchester plc pursuant to a scheme of arrangement under section 425 of the Companies Act 1985. Grantchester plc changed its name to Grantchester Holdings plc and is referred to as “GH”. At the time of that acquisition GH was the principal company of a group comprising itself and a single subsidiary, Grantchester Land Ltd, GH having acquired the latter on 28 September 2000 to form the GH group. Until 19 December 2000 GH and its subsidiary had constituted a chargeable gains group separate from GL and the members of its chargeable gains group.

GH and all members of its group, now including GL, Limitgood and Prizedome, prepared accounts for the year to 30 September 2001. Limitgood claimed to set off losses of £28,956,478 against the gains referred to at paragraph 6 above. Prizedome claimed to set off losses of £8,371,301 against gains realised during the period 19 December 2000 to 30 September 2001 and treated as accruing to Prizedome by reason of elections under section 171A on assets owned by companies which joined the GH group at the same time as Prizedome.

In their returns for the accounting period to 30 September 2002 Limitgood and Prizedome claimed to set off losses of £33,084,006 and £56,323,703 respectively against further gains realised during that period and which, by elections under section 171A, were taken as realised by Limitgood and Prizedome (in those amounts) on assets owned by companies which also joined the GH group at the same time, 19 December 2000, as they had done.

In 2002 Hammerson plc (“Hammerson”) acquired the entire share capital of GH following a public offer. As a result of Hammerson buying in excess of 29.9% of GH’s shares in the market, Hammerson was obliged under the takeover code to declare its bid unconditional when it received acceptances from over 50% of shareholders and it did so on 19 September 2002. Hammerson therefore controlled GH for a short period before GH became, by December 2002, a member of the Hammerson group for chargeable gains purposes.

Following the Hammerson takeover Limitgood prepared accounts for the twelve months to 31 December 2003 and again for the twelve months to 31 December 2004. In its return for the period to 31 December 2004 it claimed to set off £2,295,034 against gains (the “Catford gain”) elected to it by companies which joined the GH group at the same time as Limitgood.

In paragraph 12 of the Decision the Special Commissioners helpfully summarised the facts material to the issues which they were asked to determine. As so summarised the facts were that when Limitgood and Prizedome left the Anglo group on being acquired by a subsidiary of GL, losses of £113.9 million each crystallised under section 179. Shortly afterwards, but in the following accounting period, gains of £29 million were realised by other companies in the GL group and were elected to Limitgood. The GL group including Limitgood and Prizedome was acquired by GH on 19 December 2000, which then held another subsidiary; the companies in the GL group thus became members of the GH group. In their returns to 30 September 2001 Limitgood set off part of its loss against the £29 million gain and Prizedome set off losses of £8.4 million. In the year to 30 September 2002 further gains totalling £89.4 million were made by other GH subsidiaries on the disposal of assets held by these companies when previously in the GL group; those gains were elected to Prizedome and Limitgood which claimed to set off part of their losses. Following the takeover of GH by Hammerson, Limitgood claimed to set off £2.3 million of its loss against the Catford gain which had been elected to it in 2004 by companies which also joined the GH group on 19 December 2000 and which were controlled by Hammerson before becoming part of the Hammerson group.

The issues for determination by the Special Commissioners as set out in the Decision were (a) whether Limitgood was entitled to claim that gains of £29 million in the year to September 2001 were realised on assets deemed to be sold by Limitgood before the GL group was acquired by the GH group on 19 December 2000 so that Limitgood could deduct its losses under paragraphs 6(1)(a) and 7(1)(a) of schedule 7A; (b) whether Limitgood and Prizedome were entitled to claim in the case of £35.4 million gains elected to Limitgood and £64.7 million gains elected to Prizedome that the gains were realised in the years to September 2001 and 2002 on assets deemed under paragraph 7(3)(a) to be held by them immediately before they joined the GH group so that paragraph 7(1)(b) applied and paragraph 6(1)(b) required that their pre-entry losses from earlier periods must be set off against those gains, and (c) whether Limitgood was entitled to set off its losses against the Catford gain of £2.3 million realised by members of the GH group and elected to Limitgood in the period to 31 December 2004 so that paragraph 7(1)(b) again applied.

It was agreed by counsel before me that I need not go into the appeal so far as it concerned the Catford gain as, although the details of that gain had features which differed from those affecting the treatment of the other gains, my decision on the latter would govern the outcome of the appeal on the Catford gain. I therefore say no more about the Catford gain.

The basic rule in section 8(1) is that the amount to be included by a company in respect of chargeable gains in its total profits for an accounting period for corporation tax purposes is the total amount of chargeable gains accruing to the company in the period after deducting any allowable losses of the period and, so far as they have not been allowed as a deduction in any previous period, the allowable losses previously accruing to the company in earlier accounting periods. Where, however, companies are members of a chargeable gains group (as defined by section 170 to which I come later), assets may be transferred between group companies on a “no gain no loss” basis. See section 171. Prior to 1 April 2000, there was no system under which capital losses realised by one group company could reduce chargeable gains realised by another in the way that the income losses of one group company may be set off against the profits of another. Offsetting chargeable gains and allowable losses within a group could, however, be achieved by transferring assets between group companies on a no gain no loss basis prior to disposing of the assets out of the group. In this way the chargeable gains and allowable losses could be realised (and offset) within a single group company. Since 1 April 2000 section 171A has removed the need to transfer assets from one group company to another in order to offset gains and losses in this way. It is now possible for a group company that sells an asset outside the group to elect jointly with another group company that the asset shall be treated as if the former had transferred the asset to the latter before the former’s sale of the asset outside the group. In this way the gain or loss arising on the former’s disposal is treated as arising to the latter rather than to the former.

Before the enactment in 1993 of schedule 7A a group could buy a company with realised allowable losses (or owning assets that were subsequently sold at a loss) - pre-entry losses - and shelter the group’s gains by transferring to that company assets standing at a gain prior to the sale of them by the company out of the group. The group would thereby acquire the company with losses and transfer assets to it under section 171; the company would then realise the gains on the assets by selling them outside the group and by those means offset its losses against those gains. Schedule 7A was introduced to prevent this to the extent of such pre-entry losses. It did so in respect of accounting periods ending on or after 16 March 1993 (subject to certain transitional rules that are not relevant to the present case) by restricting the categories of gains against which the company was entitled to deduct such losses.

Since I am not concerned with the position where a company joins a group holding a so-called “pre-entry asset” (that is, assets pregnant with loss at the time of the company’s entry into the relevant group but which are only realised later after the company has joined the relevant group) I do not need to advert to the detailed provisions contained in schedule 7A which are concerned with pre-entry assets. I need only say that pre-entry losses affecting pre-entry assets give rise to complications which are not present in the case of realised pre-entry losses, ie losses which have been realised by the company in question before it becomes a member of the relevant group.

Before coming to the material provisions of schedule 7A, it is necessary to understand the concept of a chargeable gains group for the purposes of the pre-entry loss rules set out in the schedule. The concept is explained in section 170 which, so far as material, provides as follows:

“(1)

This section has effect for the interpretation of sections 171 to 181 except insofar as the context otherwise requires …”

Schedule 7A is given effect by section 177A and is therefore to be interpreted in accordance with section 170 unless the context otherwise requires. It is plain from a reading of the schedule that the context does not otherwise require with the result that the interpretative provisions of section 170 apply to schedule 7A.

“(2)

Except as otherwise provided -

…

(b)

subsections (3) to (6) below apply to determine whether companies form a group and, where they do, which is the principal company of the group;

…

(3)

Subject to subsections (4) to (6) below -

(a)

a company (referred to below and in sections 171 to 181 as the “principal company of the group”) and all its 75 per cent subsidiaries form a group and, if any of those subsidiaries have 75 per cent subsidiaries, the group includes them and their 75 per cent subsidiaries, and so on, but

(b)

a group does not include any company (other than the principal company of the group) that is not an effective 51 per cent subsidiary of the principal company of the group.”

The reason for subsection 3(b) is, I understand, to ensure that a subsidiary at the bottom of the ladder of a group of companies, although a 75% subsidiary of its immediate parent, is an effective 51% subsidiary of the principal company of the group. It will not follow that, because each is a 75% subsidiary of the other, the subsidiary at the bottom of the ladder is an effective 51% subsidiary of the principal company. (The expression “an effective 51% subsidiary” is explained in subsection (7).)

“(4)

A company cannot be the principal company of a group if it is itself a 75 per cent subsidiary of another company.

…

(6)

A company cannot be a member of more than one group; …

…

(10)

For the purposes of this section and sections 171 to 181, a group remains the same group so long as the same company remains the principal company of the group, and if at any time the principal company of a group becomes a member of another group, the first group and the other group shall be regarded as the same, and the question whether or not a company has ceased to be a member of a group shall be determined accordingly.”

It is common ground between the parties that GH’s acquisition of the GL group (headed by GL) on 19 December 2000 had the consequence that section 170(10) required that the former GL group and the group headed by GH as it stood from and after the takeover (ie, GH and its group as constituted immediately before the takeover, together with the members of the former GL group) be regarded as the same group and that this means, in effect, that it is to be regarded as a single, continuing group. I would merely observe that I raised with Mr Malcolm Gammie QC, who with Mr David Ewart QC appeared for HMRC, whether section 170(10) meant that the group as constituted after the takeover is to be treated as a continuation of the group that had been taken over. I understood Mr Gammie to say that, at any rate from the perspective of the companies that are taken over, the group of which they have been members continues as the same group after takeover. Indeed, some coloured diagrams produced by Mr Gammie and Mr Ewart illustrated that, by force of section 170(10), the GL group continued seamlessly as the same chargeable gains group (for the purposes of the provisions with which I am concerned) notwithstanding takeover by the GH group on 19 December 2000 and by the Hammerson group in December 2002. For convenience in argument the group was referred to on and following takeover by the GH group as the GL/GH group and, on and following takeover by the Hammerson group, as the GL/GH/Hammerson group. It was no part of the submissions of Mr Graham Aaronson QC and Mr James Henderson, appearing for the respondents, to suggest otherwise. Their observation on the operation of section 170 was that the legislation looks at the group from the top down, ie from the perspective of the principal company of the group, and that, for the purposes of the legislation, once the GL group had been taken over by GH becoming the principal company of the merged group, the GL group ceased to exist; instead, there was simply the GL/GH group (or GH/GL group as Mr Aaronson preferred to call it given his reference to the “top down” approach) which by force of section 170(10) is deemed to be the same group as the (former) GL group.

Although relevant to the facts affecting this appeal, I do not set out any parts of section 171 (which contains general provisions concerned with transfers within a group), section 171A (concerned with “notional” transfers within a group) or section 179 (concerned with the concept of a company ceasing to be a member of a group) since the fiscal impact of those provisions, so far as they affect the transactions with which this appeal is concerned, is not in dispute and has been sufficiently summarised in the statement of agreed facts set out above.

I now come to the provisions of schedule 7A material to this appeal:

“1(1) This Schedule shall have effect, in the case of a company which is or has been a member of a group of companies (“the relevant group”), in relation to any pre-entry losses of that company.

(2)

In this Schedule “pre-entry loss”, in relation to any company, means -

(a)

any allowable loss that accrued to that company at a time before it became a member of the relevant group; or

(b)

the pre-entry proportion of any allowable loss accruing to that company on the disposal of any pre-entry asset;

…”

A “pre-entry asset” is, essentially, an asset that the company in question owned when it became a member of the relevant group. The “pre-entry proportion” of any loss on any pre-entry asset is, essentially, the proportion of any loss arising on the “post-entry” disposal of a pre-entry asset that is attributable to the period of ownership of the asset prior to the company joining the relevant group. Schedule 7A contains detailed rules which deal with the identification of pre-entry assets in special circumstances and for calculating the pre-entry proportion of any loss on a pre-entry asset. As I have mentioned, these particular rules are not relevant to this appeal.

“(6)

Subject to so much of sub-paragraph (6) of paragraph 9 below as requires the groups of companies to be treated as separate groups for the purposes of that paragraph, if -

(a)

the principal company of a group of companies (“the first group”) has at any time become a member of another group (“the second group”) so that the two groups are treated as the same by virtue of subsection (10) of section 170, and

(b)

the second group, together in pursuance of that subsection with the first group, is the relevant group,

then, except where sub-paragraph (7) below applies, the members of the first group shall be treated for the purposes of this Schedule as having become members of the relevant group at that time, and not by virtue of that subsection at the times when they became members of the first group.”

This is the provision which lies at the heart of this appeal. Where it applies, it operates to delay the date when companies become “members of the relevant group” to the date that “the first group” has been acquired by the “second group”. It is material therefore to the identification of the pre-entry losses.

“(7)

This sub-paragraph applies where -

(a)

the persons who immediately before the time when the principal company of the first group became a member of the second group owned the shares comprised in the issued share capital of the principal company of the first group are the same as the persons who, immediately after that time, owned the shares comprised in the issued share capital of the principal company of the relevant group; and

(b)

the company which is the principal company of the relevant group immediately after that time -

(i)

was not the principal company of any group immediately before that time; and

(ii)

immediately after that time had assets consisting entirely, or almost entirely, of shares comprised in the issued share capital of the principal company of the first group.”

This deals with the situation where a new holding company is inserted at the head of an existing group. As paragraph 1(6) makes clear, it is an exception to the operation of that paragraph; the paragraph does not operate to alter the date of entry of members of the existing group to the group as enlarged by the introduction of the new holding company.

Continuing:

“6(1) In the calculation of the amount to be included in respect of chargeable gains in any company’s total profits for any accounting period -

(a)

if in that period there is any chargeable gain from which the whole or any part of any pre-entry loss accruing in that period is deductible in accordance with paragraph 7 below, the loss or, as the case may be, that part of it shall be deducted from that gain;

(b)

if, after all such deductions as may be made under paragraph (a) had been made, there is in that period any chargeable gain from which the whole or any part on any pre-entry loss carried forward from a previous accounting period is deductible in accordance with paragraph 7 below, the loss or, as the case may be, that part of it shall be deducted from that gain;

(c)

the total chargeable gains (if any) remaining after the making of all such deductions as may be made under paragraph (a) or (b) above shall be subject to deductions in accordance with section 8(1) in respect of any allowable losses that are not pre-entry losses; and

(d)

any pre-entry loss which has not been the subject of a deduction under paragraph (a) or (b) above (as well as any other losses falling to be carried forward under section 8(1); shall be carried forward to the following accounting period of that company.

…

7(1) A pre-entry loss that accrued to a company before it became a member of the relevant group shall be deductible from a chargeable gain accruing to that company if the gain is one accruing -

(a)

on a disposal made by that company before the date on which it became a member of the relevant group (“the entry date”);

(b)

on the disposal of an asset which was held by that company immediately before the entry date; or

(c)

on the disposal of any asset which -

(i)

was acquired by that company on or after the entry date from a person who was not a member of the relevant group at the time of the acquisition; and

(ii)

since its acquisition from that person has not been used or held for any purposes other than those of a trade which was being carried on by that company at the time immediately before the entry date and which continued to be carried on by that company until the disposal.

…

(3)

Where two or more companies become members of the relevant group at the same time and those companies were all members of the same group of companies immediately before they became members of the relevant group, then, without prejudice to paragraph 9 below -

(a)

an asset shall be treated for the purposes of sub-paragraph (1)(b) above as held, immediately before it became a member of the relevant group, by the company to which the pre-entry loss in question accrued if that company is one of those companies and the asset was in fact so held by another of those companies;

…”

These provisions identify the gains from which the pre-entry losses are deductible and the order in which they (and other losses) may be deducted. The significance of paragraph 7(3) is that it extends the rules, set out in paragraph 7(1), which determine when a pre-entry loss can be set off against particular chargeable gains for the particular accounting period, to cases in which two or more companies become members of the relevant group at the same time where those companies were members of the same group immediately before that time. This is important because, as appears from the terms of paragraph 7(1), if the gain arises on a disposal of an asset after the company with the pre-entry loss joins the relevant group, that company must, subject to an immaterial exception, have owned the asset prior to joining the group if it wishes to set the loss against the gain. Without paragraph 7(3) the rules could therefore penalise two or more companies of one group that simultaneously join another group by denying them the benefit that existed immediately before that time (and to which they were entitled) of “grouping” gains and losses by means of intra-group transfers or, from 1 April 2000, by means of an election under section 171A.

Continuing:

“9(1) This paragraph shall apply where there is more than one group of companies which would be the relevant group in relation to any company.

(2)

Where any loss has accrued on the disposal by any company of any asset, this Schedule shall not apply by reference to any group of companies in relation to any loss accruing on that disposal unless -

(a)

the group is a group in relation which that loss is a pare-entry loss by virtue of paragraph 1(2)(a) above or, if there is more than one such group, the one of which that company most recently became a member;

…

(6)

Notwithstanding that the principal company of one group (“the first group”) has become a member of another (“the second group”), those two groups shall not by virtue of section 170(10) be treated in relation to any company that is or has become a member of the second group (“the relevant company”) as the same group for the purposes of this paragraph if -

(a)

the time at which the relevant company became a member of the first group is a time in the same accounting period as that in which the principal company of the first group became a member of the second group; or

(b)

the principal company of the first group was under the control, immediately before it became a member of the second group, of a company which at that time was already a member of the second group.

…”

Under paragraph 1(1) the relevant group is, as has been seen, “a group of companies” of which the company in question “is or has been a member”. If a company has only ever been in one group, this will be the relevant group, but if it has been in more than one group, paragraph 9 identifies the group or groups to which the Schedule is to be applied. On the facts of this appeal, the only group identified by paragraph 9 is, so far as material, the merged GL/GH group. Paragraph 9(6) contains a particular provision for the case where the principal company of one group has become a member of another group. Where it applies, the merged groups are to be treated as separate groups for the purposes of paragraph 9. It is common ground that paragraph 9(6) does not apply here.

It is clear, and there is no dispute, that the respondents’ losses - the £113,921,249 loss which each realised, with effect from 26 September 2000, on quitting the Anglo group - were pre-entry losses in relation to the GL group when on 27 September 2000 they became members of that group. It is also clear - and there is equally no doubt - that from 27 September until at least 19 December 2000 (when the GL group was taken over by the GH group) Schedule 7A prevented the respondents from setting those losses against chargeable gains which were realised by other GL companies on 12 October 2000 or which might be realised subsequently.

The question raised by this appeal is whether, as the respondents successfully contended before the Special Commissioners (albeit, in view of the division of opinion between them, as the result of the chairman’s casting vote) but HMRC disputed, the respondents became entitled, as a result of the companies in the GL group (including the respondents) becoming members of the GH group on 19 December 2000 (thus giving rise to the GL/GH group), to set those losses against the chargeable gains made on the disposal of assets by other members of the GL group (as it then existed) on 12 October 2000 and subsequently between 1 October 2001 and 30 September 2002. The gains were treated as accruing to the respondents by elections under section 171A and, by force of paragraph 7(3)(a), were to be treated for the purposes of paragraph 7(1)(b) as if the assets in question had been held by the respondents immediately before they became members of the GL/GH group on 19 December 2000. The resolution of this question turns, as the parties accept and as the Special Commissioners found, on the true meaning and effect of paragraph 1(6). It turns, in particular, on whether that paragraph applies so as to treat the members of the GL group as having joined the GL/GH group on 19 December as regards the respondents’ pre-entry losses, ie the loss of £113,921,249 which accrued to each respondent prior to its entry into the GL group on 27 September 2000.

The Chairman

Having concluded, at paragraph 39, that the appeal turned on the interpretation of paragraph 1(6)(b) and, more specifically, on whether that sub-paragraph referred to “the second group” separately as “the relevant group” or both “the second group” and “the first group” and on the applicability of paragraph 9(2)(a), the Chairman went on to state, at paragraph 44, that:

“As a matter of pure language para 1(6)(b) specifically applies when the second group and the first group together are the relevant group. As a matter of language it is quite impossible to regard the words of para 1(6)(b) as applying when the second group without the first group is the relevant group. Such an interpretation would be the very reverse of the natural meaning of the words of the statute. The “second group” and “first group” clearly have the same meaning in para 1(6)(b) as in para 1(6)(a). The “second group” is a group of which “the first group” has become a member so that s170(10) applies.”

He then considered the rival contentions of the parties and, at paragraph 50, stated that:

“In deciding whether para 1(6)(b) is satisfied it is necessary to see whether the group referred to in para 1(6)(b) is the relevant group. The group referred to is “the second group, together in pursuance of [s170(10)] with the first group.” Those are the words appearing before “is the relevant group”. In effect the Revenue submission involves reading the sub-para as saying “the second group would but for the application of that subsection be the relevant group.” It is impossible to extract from the words actually used a test that involves considering the first group and the second group as separate groups. Nor is it apparent how the words can be read as asking whether before applying s170(10) the second group is the relevant group. Such an interpretation would not merely strain the language of the statute: it would give a wholly different meaning to the words used.”

He therefore rejected the Revenue approach. At paragraph 51, after stating that none of the submissions concerned with the legislative history of the Schedule was of any assistance, he continued:

“Although that sub-para [paragraph 1(6)(b)] produces a result in the present case which may not have been intended, there is nothing ambiguous about the wording used … It appears that facts such as those in these appeals were not considered by the draftsman with the result that they were not covered by the very detailed and specific provisions enacted. ”

He next considered paragraph 9 and concluded, at paragraph 61, that “…although awkwardly worded, para 9(1) does apply … and therefore para 9(2)(a) applies”. He then continued:

“The result of applying para 9(2)(a) to the facts of this case is as follows. All the losses in question accrued to the Appellants on the deemed disposal of their shares in Coalite. For the purposes of para 9(1) there is more than one group in relation to which the losses in question are pre-entry losses. Schedule 7A does not apply by reference to any group in relation to those losses unless the group is that of which the Appellant in question most recently became a member. In relation to the Catford gain that group is the Hammerson group … In relation to the other gains that group is the GH group (or the GL/GH group). The result is therefore the same as that reached on Mr Aaronson’s interpretation of para 1(6)(b).”

I take that to be a reference to the first submission of Mr Aaronson, referred to in paragraph 47 of the Decision, that paragraph 1(6)(b) was not a separate condition but “merely a consequence”. He then stated that:

“62.

Consideration of para 9(1) reinforces concerns as to the clarity in drafting of Sch 7A including para 1(6). The drafting of para 9(1) singularly fails to achieve a clear result. The wording of para 1(6)(b) is clear while apparently adding nothing to para 1(6)(a). The interpretation for which the Revenue contends no doubt conforms with the broad intention of Sch 7A, however it seems clear that the words used were not selected with the present facts in mind and cannot be interpreted as the Revenue contend without doing undue violence to the language used. Furthermore, the Revenue’s interpretation conflicts with para 9(2)(a)…”

I understand the Revenue’s contentions before the Special Commissioners - at any rate as set out in the skeleton argument on that occasion - were in substance the same as those placed before me.

Having stated that the same reasoning applied to the Catford gains, he concluded that the appeals succeeded.

In truth, as he recognised, the Chairman treated paragraph 1(6)(b) as adding nothing to the requirements of paragraph 1(6)(a) even though, at paragraph 45 of the Decision, he had accepted as correct a submission that paragraph 1(6) is structured so as to require two separate condition (set out in sub-paragraphs (a) and (b) respectively) to be fulfilled. This led to the conclusion that when the members of one group of companies become members of a second group in the circumstances set out in paragraph 1(6)(a) then, provided the combined group is the relevant group (applying paragraph 9 if necessary to determine whether this is so), the operation of section 170(10) is modified with the result that the members of the first group are to be regarded as having become members of the combined group (and therefore the relevant group) at the date when they become members of the second group rather than as at the dates when, as section 170(10) would otherwise have required, they became members of the first group. Having reached that conclusion, the Chairman assumed that any losses brought into the combined group by member companies of the first group would rank as pre-entry in relation to the combined group but would not to any extent rank as pre-entry as regards any of the gains of the first group. In effect, following the date of joining the merged group, losses in any of the companies forming the first group could only thenceforth rank as pre-entry in relation to the combined group.

Dr Avery Jones

The kernel of Dr Jones’ dissent from the Chairman’s conclusions (and therefore from the outcome of the appeal to the Special Commissioners) is to be found in the following paragraphs of the Decision:

“64[first] The problem is that the legislation recognises as the relevant group only a deemed world group because it is built on the existence of s170(10), but the operative part of para 1(6) (apart from the reference to the relevant group) operates in the real world to define the time a company joins the relevant group to be the time of the merger of the two groups, which is a non-event in the deemed world. The issue being addressed by sub-para (b) is not whether in the deemed world the combined group is the relevant group - that is the only possible relevant group in the deemed world - but whether in the real world it is the second group or the first group that is the relevant group. If the answer is the former, then in the deemed world it must be taken together with the first group in pursuance of s170(10); and if the latter, then in the deemed world it must be taken together with the second group in pursuance of s170(10).

65.

There are three reasons which seem to me to support this interpretation. If the draftsman had intended sub-para (b) to say loosely something to the effect that “if the combined group is the relevant group,” a more natural wording, following immediately from sub-para (a) would have been something like “that same group is the relevant group.” Secondly, it explains why the draftsman had put the second group first. It hardly seems likely given the style of drafting of the whole schedule that this was done for poetic effect. Thirdly, it means that sub-para (b) operates as a genuine condition. Given the context of changing the time of a company joining the combined group it does not seem likely that sub-para (b) was included merely to confirm that one is dealing with the situation of the company still being in the combined group rather than in a wholly separate group.

64[second] This reading gives the legislation the effect one would expect: if in the real world the second group is the relevant group (that is, in relation to a realised loss, the loss is realised while the company was in the first group and before it joined the second group) the legislation applies as if the members of the first group had become members of the relevant group at the time of the merger of the two groups; but if in the real world the first group is the relevant group (that is, in relation to a realised loss, the loss was realised before the company joined the first group) the legislation applies as if the members of the first group became members at the time they actually joined the first group. With regard to [the Chairman’s] suggestion (paragraph 50 above) that the Revenue’s submission involves reading the sub-para as saying “the second group would but for the application of the subsection be the relevant group,” I consider that this is exactly what the legislation says, except that it does it the other way round since the draftsman starts in the real world from the position that s170(10) does not apply, and then adds the effect of its applying by the words “together in pursuance of [s170(10)] with the first group”.”

He then enlarged on this in paragraphs 66 and 67. In paragraphs 68 to 70 he went on to say that he did not find any assistance from the rest of the Schedule, explaining in particular why (as he summarised the matter in paragraph 70) there was nothing in paragraph 9 that affected the application of paragraph 1(6) or helped to construe that paragraph. He then applied his interpretation of paragraph 1(6), as set out in paragraph 64(first), by stating:

“71.

Applying this interpretation to the facts of the main part of this case, the loss was realised before the Appellants became members of the first group (the GL group). That group is accordingly the relevant group; and the second group (the GH group) is not the relevant group (before applying s170(10)). The condition in para 1(6)(b) is not satisfied and so the time of entry of the Appellants into the relevant group is the date they became members of the first (GL) group. I would therefore decide that the losses are not available for offset against the gains made by other members of the GL group.”

In reaching his conclusion Dr Avery Jones followed his own decision in Five Oaks Properties Ltd v Revenue and Customs Commissioners [2006] STC (SCD) 769 although his reasoning in that case was in some respects different from his reasoning in the instant case.

The broad contention of HMRC is that the losses of the two respondents were pre-entry losses of the GL group, that they remained pre-entry losses of the GL/GH group and that they cannot therefore be set against the chargeable gains in question. In a very closely reasoned skeleton argument, Mr Gammie and Mr Ewart submitted as follows.

The first and fundamental question that has to be answered when applying Schedule 7A is: when did the particular company become a member of the relevant group? This is because the Schedule only has effect in relation to pre-entry losses and whether a loss is a pre-entry loss depends upon when the company with the loss became a member of the relevant group. The question has to be answered for the relevant accounting period in relation to each loss incurred (or asset pregnant with loss owned) by a company to determine whether it is a pre-entry loss and, if so, to determine the particular gain that the particular loss may reduce.

Identity of the “relevant group” to which the Schedule refers depends upon the loss that has been identified as pre-entry in respect of which the Schedule has effect. In paragraph 1(2) of the Schedule the relevant group is the group by reference to which the loss is a pre-entry loss. In most cases the group which is the relevant group for the purpose of applying the computational rules for a particular accounting period will be obvious. If, however, the company has been part of two or more groups in an accounting period and the losses are pre-entry losses of more than one of those groups, paragraph 9 comes into play. It does so, according to sub-paragraph (1) of paragraph 9, “where there is more than one group of companies which would be the relevant group in relation to any company”. Paragraph 9 does not apply to the losses claimed by the respondents in the accounting periods ending 30 September 2001 and 30 September 2002 since, so far as material, there is only one relevant group. Initially this was known as the GL group, which the respondents joined on 27 September 2000. On 19 December 2000 the GL group was acquired by the GH group and in December 2002 the GL/GH group joined the Hammerson group. On both occasions, however, section 170(10) prescribed that the GL group, the GL/GH group and the GL/GH/Hammerson group were to be treated as “the same group” (to adopt the phraseology of that provision).

To identify whether paragraph 7 of the Schedule permits the respondents to set their losses against the gains of other GL group companies in their accounting periods ended 30 September 2001 and 30 September 2002, it is necessary to identify when the respondents joined the relevant group. In the absence of any rule to the contrary, the answer is 27 September 2000 when GL’s wholly-owned subsidiary acquired the respondents from Anglo. The only relevant rule that could lead to a contrary conclusion is in paragraph 1(6) of the Schedule.

Paragraph 1(6) deals solely with the time at which a company is to be treated as joining the relevant group by reference to which a loss will be pre-entry or not. The final words of the sub-paragraph, namely “and not by virtue of that subsection [ie section 170(10)] at the times when they became members of the first group”, confirm the general rule that a company becomes a member of the relevant group when it first joins the group. Because section 170(10) treats the first and second groups as the same group, the second (GH) group’s acquisition of the first (GL) group would not affect the date of entry of the companies to the first (GL) group but for the exception provided by paragraph 1(6). Thus it is clear that 27 September 2000 is the respondents’ entry date to the GL/GH group unless paragraph 1(6) applies to give a different date.

As an exception to the general rule, paragraph 1(6) treats members of the first (GL) group as becoming members of the relevant group when the second (GH) group acquires the first (GL) group and not on their actual date of entry to the second (GH) group, but only in the circumstances specified in the sub-paragraphs (a) and (b) of that paragraph. Sub-paragraph (a) specifies that the exception applies in circumstances in which section 170(10) has operated and, in the process, identifies as two separate and distinct groups for the purposes of what follows in paragraph 1(6), “the first group” (ie the GL group) and “the second group (ie the GH group). Sub-paragraph (b) requires that “the second group” [ie the GH group] together in pursuance of that subsection [ie section 170(10)] with the first group [ie GL group] is the relevant group”. Parliament must have intended that sub-paragraph to add something to sub-paragraph (a). If Parliament meant that on any date when members of the first group should become members of the second group, so that section 170(10) applied, the members of the first group should be treated for the purposes of the Schedule as becoming members of the relevant group at that time (irrespective of their actual entry dates to the first group), Parliament could have dispensed with sub-paragraph (b). Accordingly, the issue is to understand when the second group, together with the first group, “is the relevant group” in relation to a particular loss.

The simple answer to this is that the second group (in this case the GH group) is the relevant group when the first group (in this case the GL group) is not the relevant group in relation to a particular loss (so that, in the circumstances of this case, the GL/GH group, being by force of section 170(10) the same group as the GL group is not the relevant group in relation to that particular loss). If, however, the losses are pre-entry losses of the first (GL) group, section 170(10) operates to ensure that they are and remain pre-entry losses of the merged (GL/GH) group. This is because under section 170(10) the GL/GH group is the same group as the GL group and Schedule 7A automatically applies to losses that were already pre-entry losses of the first (GL) group. In that situation the GL/GH group is the relevant group in relation to such losses (or assets): only paragraph 1(6)(a) is engaged. Paragraph 1(6)(b) is not as it adds nothing to (a).

On the other hand, without the exception in paragraph 1(6), the GL/GH group would not, where section 170(10) applies, be the relevant group in respect of such of the losses of the GL group (considered as the first group) as were not pre-entry losses of the GL group but which would be pre-entry losses of the GH group considered as the second group, ie as a distinct or separate group from the GL (and GL/GH) group. In that situation the second group (the GH group considered as a second group distinct from the merged GL/GH group, but necessarily with the addition of the GL group companies) is the relevant group in relation to those GL group losses. The GH group, as so considered, is the relevant group because, by virtue of section 170(10), the GL/GH group is not the relevant group in relation to those losses. In contrast, therefore, to the position described in the preceding paragraph, the requirements of both condition (a) and condition (b) of paragraph 1(6) are satisfied and it is only therefore in those circumstances that the exception to the general rule applies, thus ensuring that the Schedule applies to those (GL) losses (which would otherwise not be pre-entry losses) following (and by reference to) the acquisition of the first (GL) group by the second (GH) group.

Thus, the particular function of paragraph 1(6), as indicated by the condition set out in paragraph 1(6)(b), is, when applied to the current case, to bring within the scope of the Schedule those losses of the GL group that would not otherwise fall to be treated as pre-entry losses of the GL/GH group if ordinary effect were to be given to the rule contained in section 170(10). It determines whether there is by reference to the GH group (as the relevant group) a pre-entry loss in respect of which the Schedule has effect. It operates to update the time of entry to the relevant group of the companies in the first (GL) group as regards losses that are not pre-entry losses of the first (GL) group but are pre-entry losses of the second (GH) group following its acquisition of the first (GL) group. Adopting the language employed by Dr Avery Jones in his approach to paragraph 1(6) (and set out earlier) the paragraph operates in the “real world” in that, in the context of that paragraph, the first (GL) group and the second (GH) group are being considered in their “real world” capacities as separate and distinct groups whereas the merged group to which the paragraph refers is being considered in a “deemed world” sense - by force of section 170(10). Without the deeming of paragraph 1(6), section 170(10) would mean that such losses would not become pre-entry losses on the merger with the result that the Schedule would not operate in relation to those losses even though the merger combines two groups and offers the opportunity for offsetting losses of the one group against the gains of the other.

Coming to the present case, the respondents’ losses were pre-entry losses of the GL group. Accordingly, on the acquisition of GL by the GH Group on 19 December 2000, those losses remained pre-entry losses of the merged GL/GH group. There is no need to modify the usual rule identifying when the respondents joined the group; the respondents’ (pre-entry) losses cannot be set against gains, whether derived from the GL group or from the GH group. Paragraph 1(6) does not therefore affect the respondents’ date of entry into the relevant group, the GL/GH group, in relation to the losses in question. This remains unchanged on and following 27 September 2000 because those losses were pre-entry to the GL group and were therefore pre-entry to the merged GL/GH group, which is treated as the same group by virtue of section 170(10).

On behalf of the respondents Mr Aaronson accepted, as he had before the Special Commissioners, that the broad purpose of schedule 7A was to prevent losses being brought into a group and set off against group gains. He also accepted that the arrangements in this case had been contrived in order to obtain relief on that very basis. He accepted, indeed he asserted, that the legislation should be interpreted to prevent such arrangements from succeeding if and only if it is reasonably possible to do so on the language used in the schedule. What was not permissible was to distort and contort the words used in order, as he put it, to “squeeze out” a meaning that the draftsman did not intend, and could not have intended, the words to bear.

It was not possible in the present case to interpret the legislation to accord with Parliament’s overall intention to preclude such losses from being set off in this manner because the draftsman had adopted a prescriptive approach to the drafting of the Schedule rather than, as has since become the norm in taxing statutes, adopting a broad based rule directed at preventing abuse. Indeed, Parliament has since chosen to do so in this very area: see section 70 of the Finance Act 2006 inserting, among other provisions, a new section 184A into the 1992 Act and amending schedule 7A. It was plain that the draftsman of schedule 7A did not have in mind the “fact pattern” thrown up by the present case and that, had he had such pattern in mind, he might have structured the rules differently to deal with it. It was not possible, as a matter of legislative interpretation, to ascribe to paragraph 1(6) a meaning that made no sense linguistically simply in order to prevent the deduction of losses in a situation which was not in the mind of the draftsman when the legislation was framed. In short, HMRC was seeking to substitute an interpretation of paragraph 1(6) which conflicts with its plain meaning and is inconsistent with the rest of schedule 7A and which the draftsman did not and could not have intended.

The effect of paragraph 1(6), on its true construction and applying that construction to the facts of this case, is that members of what was the GL group (ie the group of which the principal company was GL) are to be treated as joining the GL/GH group on 19 December 2000. This means that the respondents’ losses, being pre-entry losses within the meaning of paragraph 1(2)(a) (“any allowable loss that accrued to that company at a time before it became a member of the relevant group”), are losses incurred before 19 December 2000 when, by force of paragraph 1(6), the respondents are to be treated for the purposes of schedule 7A as having become members of the GL/GH group. This has the consequence that those losses are deductible (1) from chargeable gains made by the respondents or treated, by elections made under section 171A, as made by them before 19 December 2000 and (2) from chargeable gains where the gains accrue on the disposal, whenever occurring, of assets held before that date by other subsidiaries in the GL group which, together with the respondents, were members of the GL group immediately before they became members of the GL/GH group on that date and which, by paragraph 7(3)(a), are to be treated for the purposes of paragraph 7(1)(b) as if they had been held by the respondents immediately before they became members of the GL/GH group on that date.

Paragraph 1(6) applies so that the members of the GL group, including the respondents, are treated as joining the GL/GH group on 19 December 2000 because, applying paragraph 1(6)(a), GL which was the principal company of the GL group (identified as “the first group” in paragraph 1(6)) became on that date a member of another group (identified in paragraph 1(6) as “the second group”) the principal company of which was GH, thereby forming the GL/GH group. Applying paragraph 1(6)(b), the second group to which that sub-paragraph refers is the merged GL/GH group. The inclusion in paragraph 1(6)(b) of the words “together in pursuance of that subsection [section 170(10)] within the first group” makes explicit what is in any event implicit, namely, that the pre-merger group of which GL was the principal company is part of the group of which GH is the principal company after the merger, namely the GL/GH group. Paragraph 1(6)(b) requires that the “second group, together …with the first group” - ie the merged group - should be “the relevant group”. In the circumstances of this case, there can be only one “relevant group”, namely the GL/GH group.

In any event, paragraph 9 provides rules to determine which is the relevant group where there is more than one group of companies which would be the relevant group in relation to any company. It operates as a filter. There is no need to consider the operation of paragraph 9 in the instant case because the GL/GH group is the relevant group and there is no other contender. But even if there were - and it is accepted that, on one view of paragraph 9(1), the paragraph can only apply where there is more than one group of companies which would be the relevant group in relation to any company - paragraph 9(2)(a) would identify the GL/GH group as the relevant group. Since it was common ground that paragraph 9(6) did not apply, and that paragraph 1(7) did not either, it followed that the GL/GH group is the entity identified in paragraph 1(6)(b) and therefore that it is the relevant group. The result is that paragraph 1(6) applies and the respondents are to be treated as having become members of the GL/GH group on 19 December 2000. That being so, their losses (ie the £113,921,249 which accrued to each respondent with effect from 26 September 2000) are necessarily pre-entry losses in relation to the GL/GH group. But so also, crucially, are the gains realised on the disposal of assets held by other members of the former GL group from which the deductions on account of those losses have been claimed.

It is not accurate to say that paragraph 1(6)(b) is redundant on the respondents’ interpretation. Paragraph 1(6)(a) looks at when a group “has at any time” joined another group while paragraph 1(6)(b) is concerned with whether the merged group “is the relevant group”. These are separate requirements in which different factual circumstances are being considered at different times. It is perfectly possible to envisage circumstances in which two groups may be treated as the same by virtue of sections 170(10) but, nevertheless, the merged group is not the “relevant group” for the purposes of paragraph 1(6)(b). In that event the paragraph does not apply.

There is nothing surprising about this result. Paragraph 47569 of HMRC’s Inland Revenue Manuals (produced for the assistance of tax inspectors and now available to the public generally) support the respondents’ interpretation of paragraph 1(6). That paragraph, so far as material, states as follows:

“There is a special rule for the case where one group takes over another group. The general rule in section 170(10) TGCA 1992… is that if the principal company of one group becomes a member of another group, the two groups are regarded as the same. Paragraph 1(6) Schedule 7A makes clear that, where the principal company of the X group becomes a member of the Y group, the members of the X group are treated as becoming members of the Y group at the time of the takeover for the purposes of Schedule 7A. The alternative interpretation based on Section 170(10), which paragraph 1(6) Schedule 7A prevents, would be that, since the X group and the Y group are treated as the same group, the X group companies are treated as joining the Y group at the time they originally joined the X group. The Schedule 7A rules accordingly restrict the deduction of losses from gains accruing after the takeover, if the losses were realised losses brought into the Y group by the X group, or the losses accrued on the disposal of assets brought into the Y group by the X group. The general effect is that where the Y group acquires company X, and X brings into the Y group its wholly owned subsidiaries, XA and XB, the loss restrictions are the same whether or not X is itself the principal company of a group…”

The Manual mentions no exceptional circumstances where the rule does not apply other than the limited exception, in paragraph 1(7), mentioned in the next paragraph of the Manual, which, it is common ground, is not in point in the present case.

HMRC’s argument seeks wrongly to link the identity of the relevant group with a particular pre-entry loss. That is to misunderstand the concept of a “relevant group” for the purposes so Schedule 7A. The concept is introduced in paragraph 1(1) as a general concept only; crucially, it is not defined by reference to the pre-entry losses of a company. Indeed if the draftsman had intended to achieve the result contended for by HMRC, he could easily have drafted the legislation to make this clear. For example, he could have added an additional requirement so that paragraph 1(6) would read as follows (with the additional requirement set out in bold):

“Subject to so much of sub-paragraph (6) of paragraph 9 below as requires groups of companies to be treated as separate groups for the purposes of that paragraph, if -

(a)

the principal company of a group of companies (“the first group”) has at any time become a member of another group (“the second group”) so that the two groups are treated as the same by virtue of sub-section (10) of section 170, and

(b)

the second group, together in pursuance of that sub-section with the first group, is the relevant group, and

(c)

the pre-entry loss in question was not a pre-entry loss in relation to the first group immediately before that time,

then, except where sub-paragraph (7) below applies, members of the first group shall be treated for the purposes of this Schedule as having become members of the relevant group at that time, and not by virtue of that sub-section at the times when they became members of the first group.”

To construe the paragraph in this way is to ascribe to it a meaning contrary to what it actually says. The facts of this case should not be permitted to distort the clear meaning and effect of the paragraph. It does not contain a test in relation to the status of the historic first group. On the contrary, paragraph 1(6)(b) focuses on the enlarged second group; there is no pre-entry loss test in paragraph 1(6); instead, the paragraph looks at companies and groups.

Counsel enlarged on their respective arguments in the course of their oral addresses to me, but I have, I hope, captured the essence of the very cogent submissions which each advanced.

The correct approach to statutory construction in the field of taxing statutes is, in the words of Lord Nicholls delivering the unanimous opinion of the House of Lords in Barclays Mercantile Business Finance Ltd v Mawson (Inspector of Taxes) [2004] UKHL 51 at [36], and [2005] STC 1 at [36], “first, to decide, on a purposive construction, exactly what transaction will answer to the statutory description and, secondly, to decide whether the transaction in question does so.” Lord Nicholls then quoted with approval the following passage from the judgment of Ribeiro PJ in Collector of Stamp Revenue v Arrowtown Assets Ltd [2003] HKCFA 46 at [35], (2004) 6 ITLR 454 at [35]:

“[T]he driving principle in the Ramsay line of cases continues to involve a general rule of statutory construction and an unblinkered approach to the analysis of the facts. The ultimate question is whether the relevant statutory provisions, construed purposively, were intended to apply to the transaction, viewed realistically.”

With that guidance in mind, I approach the issue of statutory construction which arises on this appeal.

As the heading to Schedule 7A indicates, and as paragraphs 1(1) and 1(2) make clear, the Schedule is concerned with what it terms “pre-entry losses” and the relationship of a company having such losses to a group of companies - termed “the relevant group” - of which the company in question is or has been a member. The object of the Schedule is to restrict the extent to which such losses may be applied to reduce chargeable gains accruing, or made available, to the company. This is achieved in paragraph 7 by tightly defining the chargeable gains from which such losses may be deducted. Broadly stated, the gains must derive from an asset which was disposed of by the company in question before the date of the company’s entry to the relevant group or was held by that company immediately before such entry date or was acquired by that company on or after such entry date from a person outside the group at the time of the asset’s acquisition. The prohibition, again stated broadly, is against allowing the losses to reduce the chargeable gains (on the disposal of assets transferred or “elected” to that company) of any members of the group which the company has joined. Paragraph 9 exists to identify which group is the relevant group where more than one group would qualify.

Because, with the passage of time, the composition of a group of companies may, and frequently does, change and the group itself may cease to be identifiable as a distinct group when it is acquired by another company or group, the legislation takes care to ensure that a relevant group, once identified as such, does not cease to be the relevant group merely because (a) other companies may join or leave it or (b) the group is acquired by another group. That requires two things: (a) a clear definition of what is meant by “a group of companies”, a matter dealt with by section 170(3) to (6), and (b) a mechanism to ensure that the scope of the restriction against deducibility is not undermined by alterations in the make-up of the relevant group. This is achieved by section 170(10) providing (a) that “a group remains the same group so long as the same company remains the principal company of the group” and (b) that “if at any time the principal company of a group becomes a member of another group, the first group and the other group shall be regarded as the same”. The subsection then states “and the question whether or not a company has ceased to be a member of a group shall be determined accordingly”. The effect of this provision, so far as it is made to apply to Schedule 7A (as by section 177A it is), is therefore to ensure that once a loss has been identified as pre-entry in respect of a group of companies identified as “the relevant group” it remains a pre-entry loss in respect of that group, notwithstanding changes in the composition of the group or the take-over of the group by another group.

Unless modified, however, section 170(10) would operate to enable losses realised by a company in one group (“the first group”) to be set against gains realised by members of another group (“the second group”) where the second group has later taken over the first group. This is because the members of the first group would, by force of section 170(10), be treated as having joined the merged group not at the time of the merger but at the time that they each became members of the first group. The losses in question would not therefore be pre-entry in relation to the merged group with the result that, as regards those losses, the merged group would not qualify as “the relevant group” and the losses would not be subject to any restrictions against deductibility.

It is to avoid this consequence that, in my judgment, paragraph 1(6) was enacted. For the effect of the paragraph, where it applies, is to treat the members of the first group as having joined the merged group at the time of the merger “and not by virtue of that subsection [ie section 170(10)] at the times when they became members of the first group.” The consequence of so doing is to treat the losses of members of the first group as pre-entry to the merged group and thus to subject them to the operation of the Schedule, ie as subject to the restrictions on deductibility set out in paragraphs 6 and 7.

The question that has arisen is whether, as HMRC contend, the operation of paragraph 1(6), where it applies, is confined to losses of members of the first group which are realised by those members while members of that group and, as such, are pre-entry to the merged group that they have joined, ie the second group together with the first group, or whether, as the respondents contend, it has a wider effect and is apt also to catch losses of any member of the first group which were realised prior to that member joining the first group and which, unaffected by the operation of paragraph 1(6), are pre-entry losses of that member in relation to the first group. The answer to this turns, as is common ground, on a true understanding of paragraph 1(6) and, in particular, of paragraph 1(6)(b).

If the respondents’ understanding of how the paragraph functions is correct and it operates to catch losses of any member of the first group which that member has realised prior to joining the first group, it is worth observing the remarkable result in this case. It is a result from which the respondents do not shy. It is that whereas until 19 December 2000 the respondents’ realised losses (each of £113.9 million) were not deductible from gains realised by members of the GL group on 12 October 2000 or subsequently (because the losses were pre-entry in relation to the GL group and the deductibility restrictions therefore applied), nevertheless by the simple expedient of the takeover of the GL group by the GH group on 19 December 2000 those losses ceased to be subject to those restrictions because the merged (GL/GH) group became the relevant group so that, by force of paragraph 1(6), the respondents and the members of the GL group were thenceforth all treated, without discrimination between them in any respect, as having joined the merged (GL/GH) group on the date of merger. The result of this, if it is correct, is that any and all realised losses of members of the former GL group (as it existed immediately prior to the merger and including therefore the respondents) became pre-entry losses in relation to the merged group and available for set-off against any gains derived from disposals by any members of the former GL group. All that is required to achieve this beneficial result is a group, which need consist of no more than two companies, to take over the GL group. (A takeover by a single company will not be sufficient: see paragraph 1(7).) The facts of this case illustrate how contrived the take-over may be in order to achieve this result. The question is whether, on a true understanding of paragraph 1(6), there is no escape from it.

Paragraph 1(6) requires two separate conditions to be fulfilled if it is to apply and the operation of section 170(10) is to be modified. The first, set out in sub-paragraph (a), is that there has at some time been a takeover of the first group by the second group so that the two groups are treated as the same by virtue of section 170(10). The second, set out in sub-paragraph (b), is that “the second group, together in pursuance of section 170(10) with the first group, is the relevant group.” The assumption underlying both sub-paragraphs is that section 170(10) operates so that the two groups are, in the words of that subsection, “regarded as the same”. Sub-paragraph (b) is intended, in my view, to add something to sub-paragraph (a): it is not enough that one group has been acquired by another in the circumstances set out in sub-paragraph (a). The addition, required by sub-paragraph (b), is that “the second group, together in pursuance of that subsection [section 170(10)] with the first group, is the relevant group”. But the relevant group in relation to what?

It is to be noted that the condition to be fulfilled by sub-paragraph (b) is that it is “the second group” that is to be the relevant group. It is not that the first group is to be that group. In adding the words “together in pursuance of that subsection with the first group” the draftsman is doing no more, in my view, than acknowledging – in line with the assumption that underlies the paragraph – that as a consequence of the operation of section 170(10) the first group and the second group are, following the takeover of the former by the latter, the same group. The effect of the paragraph is to negate the operation of that subsection so that, as regards the particular losses which are in point, the members of the first group are to be treated as having joined the relevant group at the time of the merger and not before.

What then are the losses to which the paragraph is directed? In my judgment they are losses which are pre-entry in relation to the second group; they are not the losses which are pre-entry in relation to the first group. I reach that conclusion because, if it were the latter, there would be no need to disapply the operation of section 170(10); paragraph 1(6) would add nothing to the scheme of the Schedule. It is precisely because, as regards losses which have accrued to members of a group while members of that group, there is a need, if the aim of the Schedule – to subject pre-entry losses to restrictions on set-off – is to be achieved, to disapply section 170(10) that, in my judgment, paragraph 1(6) was enacted. So regarded, it operates to put losses accruing to companies in a group which is subsequently taken over by another group on the same footing as losses accruing to a single company which is subsequently taken over by a group. That being, as I see it, the purpose of the provision, I see no reason, unless compelled by the words used to do so, to construe it as having an effect which goes beyond that purpose. I consider that the purpose can be achieved – and the surprising results avoided which I have described at paragraph 60 above - by construing the reference to “the relevant group” in sub-paragraph (b) as confined to losses of the acquired (the first) group which are not pre-entry losses in relation to that group immediately before its acquisition by the acquiring (second) group.

It follows that in reaching this conclusion I am largely in agreement with the submissions addressed to me by Mr Gammie and Mr Ewart, with their emphasis on the underlying purpose of the legislation, and I reject the more literal approach of Mr Aaronson and Mr Henderson. It follows too that I agree with the conclusion of Dr Avery Jones although it will be seen that my reasons for doing so differ in a number of respects from those which appealed to him.

In particular, I am not assisted by references to the first group and to the second group as being in the “real world” or to the relevant group, where referred to in paragraph 1(6), as being in the “deemed world” as Dr Avery Jones has done in his part of the Decision, and as Mr Gammie did in the course of his submissions. As Mr Aaronson pointed out, and I agree, the various concepts dealt with by the paragraph, namely “group of companies”, “principal company”, “relevant group” and so forth are statutory in nature. It does not assist an understanding of the paragraph to ascribe to them attributes of actual or assumed reality.

I am not troubled by the fact that the passage from the Inland Revenue Manuals to which Mr Aaronson referred is worded as it is. The particular statement in the Manuals is accurate as far as it goes. It does not purport to deal with the case where the relevant losses are pre-entry in relation to the X group (to take up the terminology of the particular passage). The most that can be said is that the passage does not differentiate between losses of the X group which are pre-entry to that group and those which are not.

Nor am I troubled by the fact that my construction of paragraph 1(6) could have been spelled out in the manner suggested by Mr Aaronson and Mr Henderson in their written skeleton argument and which I have mentioned in paragraph 51 above. Provided my interpretation of the provision is one which it is open to the court to reach, I do not think that it matters that the same result could more felicitously have been reached, and all doubts put fully to rest, if the provision had spelled it out more fully.

Nor does it assist to speculate whether, in enacting the Schedule, Parliament had in mind the circumstances of a case - or fact pattern as Mr Aaronson described it - such as are here present. The courts have frequently to apply a statutory provision to circumstances which Parliament either did not, or arguably did not, have in mind at the time of enactment. The real question is whether, given what I accept is a prescriptive approach to the drafting of the Schedule, this construction of paragraph 1(6) is one that it is open to me, bearing in mind the guidance contained in Barclays Mercantile Business Finance. For the reasons explained, I consider that it is.

I shall therefore allow the appeal and, as a result, dismiss the respondents’ appeals to the Special Commissioners.