England and Wales Court of Appeal (Civil Division) Decisions

IN THE SUPREME COURT OF JUDICATURE QBENI/1999/0292

(MR JUSTICE LIGHTMAN )

Royal Courts of Justice

Strand

London WC2

Friday, 30 July 1999

B e f o r e:

LORD JUSTICE PETER GIBSON

LORD JUSTICE PILL

LORD JUSTICE CHADWICK

- - - - - -

DUNLOP INTERNATIONAL AG

(FORMERLY DUNLOP INTERNATIONAL LTD)

Appellants

- v -

PARDOE (HM Inspector of Taxes)

Respondent

- - - - - -

(Handed Down Transcript of

Smith Bernal Reporting Limited, 180 Fleet Street,

London EC4A 2HD

Tel: 0171 421 4040

Official Shorthand Writers to the Court)

- - - - - -

MR DAVID MILNE QC & MISS ELIZABETH WILSON (Instructed by Messrs Eversheds, London, EC4V 4JL) appeared on behalf of the Appellant

MR MICHAEL FURNESS (Instructed by Solicitors of the Inland Revenue, Somerset House, London, WC2R 1LB) appeared on behalf of the Respondent

- - - - - -

J U D G M E N T

(As approved by the Court )

- - - - - -

©Crown Copyright

LORD JUSTICE CHADWICK: This is an appeal against the order made on 5 March 1998 by Mr Justice Lightman dismissing an appeal under section 56A(1) of the Taxes Management Act 1970 by Dunlop International AG (“DIAG”) from the decision of the Commissioners for the special purposes of the Income Tax Acts dated 7 January 1997. The issue on this appeal, as it was both before the special commissioners and the judge, is whether a charge to corporation tax arose, by virtue of the provisions in sections 278 (3) of the Income and Corporation Taxes Act 1970 on the transfer of residence of DIAG from the United Kingdom to Switzerland on 31 May 1978, in respect of a disposal of an asset to DIAG by a subsidiary pursuant to an agreement made on 16 March 1978. The fact that this dispute has taken so extraordinarily long to be litigated has not been explained and is not a matter for which either side reproaches the other.

The facts, which have not been in dispute, are set out at paragraphs 4 - 8 in the decision of the special commissioners:

4. At all relevant times prior to 31 May 1978 DIAG was a company incorporated with limited liability resident in the United Kingdom carrying on business as an investment holding company with investments in UK and overseas companies. During this time it was the principal company of a group of companies (“the International Group”) for the purposes of section 272 ICTA 1970 that included Dunlop Plantations Limited (“Plantations”). Until some time between 22 December 1977 and 4 April 1978 the International Group also included four other resident group companies.

5. Prior to 19 October 1973 Plantations owned 80% of the issued shared capital in Dunlop Estates Berhad (“Berhad”), a Malaysian company. On 19 October 1973 Moorgate Industrials Limited (“Moorgate”), a company resident in the UK, acquired from Plantations the shares in Berhad owned by Plantations in consideration for an issue of shares in Moorgate to Plantations. As a result Moorgate became from 19 October 1973 a member of the International Group as a 99.99% subsidiary of Plantations. Moorgate immediately sold part of its newly acquired shareholding in Berhad. The market value of its remaining 51% shareholding in Berhad (the “Berhad shares”) as at 19 October 1973 has been agreed with the Shares Valuation Division of the Inland Revenue as £10,981,307.

6. As a result of the transfer of the Berhad shares and the issue of shares in Moorgate to Plantations the International Group included in a vertical chain the three group companies for the purposes of section 272 of (i) DIAG, (ii) DIAG’s 100% subsidiary Plantations and (iii) Plantations 99.99% subsidiary Moorgate.

7. Under an agreement dated 16 March 1978 Moorgate contracted to sell the Berhad shares to DIAG for a consideration of £17,390,925. As DIAG and Moorgate were both members of the International Group section 273 ICTA 1970 applied to the sale so as to give rise to a no gain and no loss situation. As a result of this sale of the Berhad shares they became directly held by DIAG and were no longer held by Moorgate at the end of the chain of ownership of the three group companies DIAG, Plantations and Moorgate.

8. On 31 May 1978, in accordance with Treasury consent under section 482 ICTA 1970 and Bank of England exchange control consent, DIAG ceased to be resident in the United Kingdom. . . .

The relevant statutory provisions, and their legislative history (so far as material), may be summarised as follows:

(1) The Finance Act 1965, section 46(1), introduced corporation tax as a charge on the profits of companies. For the purposes of corporation tax “profits” meant income and chargeable gains; and “chargeable gains” had the meaning given to that expression in Part III of the Act. Chargeable gains were gains accruing, after 6 April 1965, on the disposal of an asset by its owner. Section 22(4) FA 1965 provided that, in the circumstances in which an asset was acquired otherwise than by way of a bargain made at arm’s length, a person’s acquisition of an asset and the disposal of it to him should, for the purposes of Part III, be deemed to be for a consideration equal to the market value of the asset.

(2) Section 55(5) FA 1965 required that, for the purposes of corporation tax, the provisions of the Act relating to capital gains tax should have effect in relation to groups of companies subject to Part I of schedule 13. Paragraph 2(1) in schedule 13 to FA 1965 - later to become section 273(1) ICTA 1970 - made special provision for the transfer of assets between companies within the same group. Such transfers were deemed to take place at a consideration which gave rise to no gain and no loss in the transferor company. The effect was that the transferee company acquired the asset for a consideration reflecting an historic acquisition value by the transferor company or one of its predecessors; so incorporating a latent liability to tax on any chargeable gain on a subsequent disposal.

(3) The “no gain - no loss” provisions in paragraph 2(1) in schedule 13 to FA 1965 were intended to postpone the payment of corporation tax on gains accruing in respect of an asset while it was held within a group until such time as the asset was transferred out of the group. But the provisions gave rise to an opportunity to postpone that time indefinitely; and so, in effect, to avoid corporation tax on gains altogether. The avoidance scheme (sometimes known as “the envelope scheme”) was described by Mr Justice Millett in NAP Holdings UK Ltd v Whittles [1992] STC 59, at page 70f-h:

"A company (P in our example), owning an asset such as a factory which had appreciated in value (from £x to £y), wished to sell it and realise the gain without incurring a charge to tax. P therefore incorporated a new subsidiary (T) and transferred the factory to it in exchange for the issue of shares in T. P then sold T (the factory in its ‘envelope’) outside the group for £y. Neither P’s transfer of the factory to T nor P’s subsequent disposal of T gave rise to a charge to tax; the former because P’s transfer was at a consideration (£x) which produced neither a gain nor a loss (s 273 [formerly para 2(1) schedule 13 FA 1965]), the latter because P’s acquisition of T’s shares was at market value (£y). The gain was realised by a sale out of the group but tax on the gain was postponed until such time, if ever, as the purchaser chose to remove the factory from its ‘envelope’."

(4) Paragraph 18(2) in schedule 12 to FA 1968 - later to become section 278(3) ICTA 1970 - was enacted in order to counteract avoidance of tax through use of the envelope scheme. As Mr Justice Millett explained in NAP Holdings UK Ltd v Whittles , at page 70h-j:

"It has effect whenever a company (T) acquires an asset from another member of the same group (P) and within six years thereafter ceases to be a member of the group. It treats T as if immediately after acquiring the asset from P (for a consideration fixed by s 273 at £x) it had sold and immediately reacquired the asset at market value (£y). The result of the notional transaction is retrospectively to make T chargeable to tax on the gain at the date of the intra-group transaction. This does not so much counteract the effect of s 273 (for T is made liable instead of P) but the indefinite postponement of T’s liability."

(5) The relevant provisions in FA 1965 and FA 1968 were repealed and re-enacted as Part XI of ICTA 1970. Provisions relating to chargeable gains were contained in Chapter II of that part. In particular, sections 272 - 279 related to transfers of assets and shares within a group of companies. Section 272 ICTA 1970 was in these terms, so far as material:

"272(1) For the purposes of this and the following sections of this Chapter -

(a) references to a company ... apply only to a company... which is resident in the United Kingdom;

(b) a principal company, and all its 75 per cent subsidiaries form a group, and where a principal company is a member of a group as being itself a 75 per cent subsidiary that group shall comprise all its 75 per cent subsidiaries.

(c) “principal company” means a company of which another company is a 75 per cent subsidiary;

(d) . . .

(e) “group” and “subsidiary” shall be construed with any necessary modifications where applied to a company incorporated under the law of a country outside the United Kingdom.

(2) . . .

(3) For the purposes referred to in subsection (1) above a group remains the same group so long as the same company remains the principal company of the group, . . ."

Section 273(1) replaced paragraph 2(1) in schedule 13 to FA 1965:

"273(1) Notwithstanding any provision in Part III of the Finance Act 1965 fixing the amount of the consideration deemed to be received on a disposal or given on an acquisition, where a member of a group of companies disposes of an asset to another member of the group, both members shall . . . be treated, so far as relates to corporation tax on chargeable gains, as if the asset acquired by the member to whom the disposal is made were acquired for a consideration of such amount as would secure that on the other’s disposal neither a gain nor a loss would accrue to that other; but where it is assumed for any purpose that a member of a group of companies has sold or acquired an asset, it shall be assumed also that it was not a sale to or an acquisition from another member of that group."

Section 278(3) ICTA 1970 replaced paragraph 18(2) in schedule 12 to FA 1968:

"278(3) If, when the chargeable company ceases to be a member of the group, the chargeable company, or an associated company also leaving the group, owns otherwise than as trading stock -

(a) [Any asset which the chargeable company acquired from another company which was at the time of the acquisition a member of that group of companies] or,

(b) . . .

the chargeable company shall be treated for all the purposes of Part III of the Finance Act 1965 as if immediately after its acquisition of the asset it had sold, and immediately re-acquired, the asset at market value at that time."

Section 278(1) defined the circumstances in which section 278(3) applied:

"278(1) If a company (in this section called the chargeable company) ceases to be a member of a group of companies, this section shall have effect as respects any asset which the chargeable company acquired from another company which was at the time of the acquisition a member of that group of companies, but only if the time of acquisition fell -

(a) on or after 6 April 1965 and

(b) within the period of six years ending with the time when the company ceases to be a member of the group;

and references in this section to a company ceasing to be a member of a group of companies do not apply to cases where the company ceases to be a member of a group by being wound up or dissolved or in consequence of another member of the group being wound up or dissolved."

But section 278(1) did not apply where it was excluded by section 278(2):

"278(2) Where two or more associated companies cease to be members of the group at the same time, sub-section (1) shall not have effect as respects an acquisition by one from another of those associated companies."

In that context “associated companies” had the meaning given by section 278(4)(a):

"278(4) For the purposes of this section - (a) two or more companies are associated companies, if, by themselves, they would form a group of companies."

The effect of section 273(1) ICTA 1970 was that, on the sale on 16 March 1978 of the Berhad shares by Moorgate, DIAG was treated as if it had acquired that asset for a consideration of such amount as would secure that neither a gain nor a loss would accrue to Moorgate on Moorgate’s disposal. In the circumstances that it has been agreed that the cost to Moorgate of the acquisition of the Berhad shares on 19 October 1973 should be treated as £10,981,307, prima facie the acquisition cost to DIAG on 16 March 1978 was deemed to be that amount. But, if section 278(3) is applicable, DIAG must also be treated as if, immediately after its acquisition of the Berhad shares on 16 March 1978, it had sold and immediately re-acquired the Berhad shares at market value at that time. If, (which is not a matter for decision) the market value of the Berhad shares on 16 March 1978 was £17,390,925, then that deemed disposal and immediate re-acquisition would appear, again prima facie , to give rise to a gain of £6,409,618 by DIAG.

For section 278(3) ICTA 1970 to be applicable three conditions must be satisfied: (i) that DIAG ceased to be a member of the International Group upon the transfer of its residence out of the United Kingdom on 31 May 1978; (ii) that, notwithstanding the transfer of residence, DIAG was, at the time when it ceased to be a member of the International Group, a “chargeable company” for the purposes of section 278(3); and (iii) that, notwithstanding that at the same time as DIAG ceased to be a member of the International Group Moorgate also ceased to be a member of that group (because the International Group ceased to exist), section 278(2) does not prevent sub-sections (1) and (3) of that section from having effect.

Mr Milne QC, on behalf of DIAG, submitted before this court - as he had before the special commissioners and before Mr Justice Lightman - that none of the three conditions which I have identified were satisfied in the present case. It is fair to say, first, that his argument on condition (iii) formed his principal submission both in this court and below (at least, before the special commissioners); and, secondly, that his enthusiasm for the argument on condition (i) was muted at the outset and waned as the argument was developed. In addressing his submissions in the order in which I have identified the three conditions, I do not intend to suggest that that reflects the degree of conviction with which they were advanced.

The question whether DIAG ceased to be a member of the International Group upon the transfer of its residence out of the United Kingdom on 31 May 1978 turns on the provisions in section 272(1) and (3) ICTA 1970. The effect of those provisions may be summarised as follows: (i) for so long as DIAG remained resident in the United Kingdom, DIAG was a “company” within the meaning of section 272(1)(a); and was a “principal company” within the meaning of section 272(1)(c) - because Plantations and Moorgate were its 75 per cent subsidiaries; (ii) for so long as DIAG was a “company” and so a “principal company”, DIAG, Plantations and Moorgate comprised a group (the International Group) for the purposes of section 272(1)(b); (iii) upon DIAG ceasing to be resident in the United Kingdom it ceased to be a “company” within section 272(1)(a) and so ceased to be a “principal company” within section 272(1)(c); (iv) accordingly, as from 31 May 1978, the International Group ceased to exist - see section 272(1)(b) and section 271(3) - and so DIAG (and also Plantations and Moorgate) ceased to be a member of that group.

Mr Milne submitted that section 278(1) could not have been intended to apply in circumstances in which a company ceased to be a member of a group because the group ceased to exist - or (at the least) where a company which had been the principal company left its own group with the result that the group thereby ceased to exist. He submitted that Parliament could not have intended that section 278(3) would have effect where, in a two company group, the subsidiary (S) is sold within six years of the transfer by S of an asset to the principal company (P) - that is to say, the obverse of the “envelope scheme” - notwithstanding that, in those circumstances, (i) the group would cease to exist on the sale of S, (ii) P and S would both cease to be members of the group, and (iii) section 278(3), read literally, would then require P to be treated as if, immediately following the transfer, P had disposed of the asset and immediately reacquired it at market value. He pointed out, correctly, that - if that were what Parliament had intended - the effect could be avoided by the simple expedient of P forming or acquiring a new subsidiary (S2) immediately before the sale of S; for, in those circumstances, P and S2 would continue to be members of the group, notwithstanding the sale of S, and section 278(3) could have no application.

In support of that submission Mr Milne drew our attention to an Inland Revenue statement of practice (IRPR 17.5.85). In answer to a request for confirmation that the Revenue would not seek to apply section 278 ICTA 1970 to assets transferred to a holding company of a group consisting of that company and one subsidiary when the subsidiary leaves the group:

"The Revenue confirmed that they would not normally seek to apply Sec 278, except where the parent company emigrated, thereby causing the group to break up."

It may be noted, of course, that the statement of practice is qualified so as to ensure that it does not extend to the present case; but Mr Milne relies upon it as an indication that the Revenue has recognised that, in the more usual two company group case - where the group ceases to exist on the sale off of the subsidiary, section 278(3) should not be applied. He reminds us of the observations of Lord Bridge in Wicks & Johnson v Firth [1983] STC 25, at page 29, that:

". . . in choosing between competing constructions of a taxing provision it is legitimate . . . to incline against a construction which the Revenue are unwilling to apply in its full rigour, but feel they must mitigate by way of extra statutory concession, recognising, presumably, that in some cases their construction would operate to produce a result which Parliament can hardly have intended."

In my view the answer to Mr Milne’s submissions on this point was given by Mr Justice Lightman in a passage in his judgment, reported at [1998] STC 459, at page 471, which I gratefully adopt:

"The concession is plainly explicable as recognising the reality in the case of a two company group where the subsidiary transfers an asset to the principal company before leaving the group that the economic ownership of the asset before and after the transfer to the principal company is the same, namely in the ownership of the principal company; and that the concession merely saves the principal company the expense and bother of creating an additional subsidiary before disposing of the existing subsidiary in order to avoid triggering an immediate tax liability. This explanation is reinforced by the consideration that the Revenue excepted from the concession the situation where the principal company emigrated.

But the conclusive answer is that there is no ambiguity in the definitions of ‘principal company’ or ‘group’. Section 272(1)(b) and (c) unambiguously define a principal company as a company of which another company is a 75% subsidiary and require both a principal company and at least one 75% subsidiary to form a group. . . . "

Mr Milne, in criticising that passage, suggested that the judge had misunderstood the purpose of the concession - which was, he submitted, to recognise that (on a literal construction of section 278) a tax liability would arise in circumstances which had not been foreseen or intended. In my view that criticism is misplaced. The judge pointed out - as is obvious - that the problem can readily be avoided by the formation or acquisition of a new subsidiary; but he had well in mind - as his observations make clear - that there was no fiscal purpose in negating (by section 278(3) ICTA 1970) the postponement of a charge to tax (effected by section 273(1) ICTA 1970) on gains in respect of an asset which would remain in the ownership of the principal company. The charge on gains would crystallise in due course, when the principal company disposed of the asset.

In my view it is plain that, on DIAG ceasing to be resident in the United Kingdom, it ceased to be a “company” - and so ceased to be a “principal company” for the purposes of section 272 ICTA 1970. Accordingly the International Group, of which DIAG had been the principal company, ceased to exist - section 272(1)(b) and (3). It must follow that DIAG ceased to be member of that group. There is nothing in the provisions of section 272 which lends support to the submission that a company can continue to be a member of the group of which it was formerly the principal company when that group has ceased to exist. The condition which I have identified as condition (i) is satisfied in the present case.

Section 278(3) ICTA 1970 applies “when the chargeable company ceases to be a member of the group”. Those words reflect the opening words in section 278(1): “If a company (in this section called the chargeable company) ceases to be a member of a group of companies”. In that context “company” means a company which is resident in the United Kingdom - section 272(1). As the judge pointed out, at [1998] STC 459, 472h-473a, the opening words in sections 278(1) and (3) are a convenient shorthand for a condition which, fully expressed, would read:

"If a company resident in the United Kingdom which is a member of a group (in this section called the chargeable company) ceases to be a member of that group of companies"

In advancing his submissions on this point Mr Milne started from the position that the reason why DIAG ceased to be a member of a group of companies on the occasion of it ceasing to be resident in the United Kingdom was that, on ceasing to be resident in the United Kingdom, DIAG ceased to be a company for the purposes of section 278. It is that proposition that underlies the conclusion that condition (i) is satisfied. Founding his argument on that proposition, he then submitted that it was necessary to construe sections 278(1) and (3) in the light of the circumstances as they existed immediately after the company had ceased to be a member of the group. By that time the company had ceased to be a company for the purposes of section 278; and so, it is said, section 278(1) and (3) could not apply to it.

The judge described the argument as “subtle”; but he rejected it as misconceived. He said this, at page 473a-b in the report of his judgment:

"Section 278 attaches the label of the ‘chargeable company’ (and fiscal consequences) to a company which, having once enjoyed the necessary residence and group membership, loses that membership for any reason. The fact that the occasion for the loss of membership (namely loss of residence) also at the same time occasions loss of qualification as a ‘company’ for purposes of the statutory scheme does not detach the label of chargeable company once so attached. This appears to me to be the plain meaning of the statutory language, and this conclusion is reinforced by the consideration that there can be no sensible reason why the legislature can have intended to distinguish, from all other companies ceasing to be members of a group, a migrating company so as to confer on a migrating company a holiday from accounting for the deferred tax charge for capital gains tax on assets removed from the group."

In my view the judge was right to reject the argument on the ground which he gave - namely that it offends the plain meaning of the statutory language. There is no reason why sections 278(1) and (3) should be construed in the light of the circumstances as they exist immediately after the company has ceased to be a member of the group. The phrase which Parliament has chosen is not “if a company has ceased to be a member”; the phrase is “if a company ceases to be a member”. The relevant question is whether the company (described in both sections 278(1) and (3) as the chargeable company) was a “company” for the purposes of section 272(1) and the other sections in Chapter II of Part XI of ICTA 1970 at the time when it ceased to be a member of the group; not at the time after it had ceased to be a member.

Mr Milne submitted that that question was meaningless in the context of a case, like the present, where the company ceased to be a company for the purposes of section 278 at the same time as it ceased to be a member of the group. The preliminary question was whether the relevant time - that is to say, the time when the company ceased to be a member of the group - was immediately before or immediately after the moment at which that event took place. I am not persuaded that the construction of section 278(1) and (3) does turn on so fine a point - interesting though it might be to a student of philosophy. But I am satisfied that, if the section was intended to apply at all to a case where a company ceases to be a member of a group on the occasion of it ceasing to be a company, there can be only one answer to the question “was DIAG a company when it ceased to be a member of the International Group?”. Plainly, it was.

Mr Milne submits that there is no reason to think that Parliament intended that section 278 should apply to a case where a company ceases to be a member of a group on the occasion of it ceasing to be a company. But it is plain that the draftsman, at least, did address his mind to that point. Section 278(1) provides, expressly, that references in section 278 to a company ceasing to be a member of a group of companies do not apply to cases where a company ceases to be a member of a group by being dissolved. That is, clearly, a case where a company ceases to be a member of a group on the occasion of it ceasing to be a company. Given the express exclusion of that particular case there is good reason to think that Parliament did intend that section 278 should apply to other cases -of which migration is an obvious example - where a company ceases to be a member of a group on the occasion of it ceasing to be a company.

Mr Milne sought to persuade us that there was no need to make provision, in section 278 ICTA 1970, for a case where the company holding the relevant asset (in respect of which tax on gains had been postponed by section 273(1)) ceased to have residence in the United Kingdom. He pointed to the need for Treasury consent, under section 482(1) ICTA 1970, and to the need for exchange control consent from the Bank of England. In my view the need for those consents - which may have been granted or withheld on grounds which went beyond considerations as to the liability to tax of the applicant - gives no guidance on the question whether section 278 was intended to apply where a company ceased to be a member of a group on the occasion of it becoming non-resident. That question must turn on the proper construction of the sections in Chapter II of Part XI of ICTA 1970.

The company could not cease to be a member of a group of companies (within the meaning of section 272(1)(b)) unless, at the time immediately before it ceased to be a member of that group, it was a member of that group. It could not be a member of the group unless, immediately before it ceased to be a member of that group, it was a “company” for the purposes of section 278(1) and (3). If section 278 was intended to apply in circumstances where a company ceases to be a member of a group on the occasion of it ceasing to be a company - which, for the reasons set out, I find to be the case - then section 278(1) and (3) must require the question “was the company a company when it ceased to be a member of the group” to be answered by reference to the circumstances immediately before it ceased to be a member. Otherwise the intention is frustrated.

For those reasons I am satisfied that the condition which I have identified as condition (ii) is satisfied.

Section 278(2) removed a case from section 278(1) and (3), in respect of an acquisition by one associated company from another, where those companies ceased to be members of the group at the same time. Companies were associated companies for that purpose “if, by themselves, they would form a group of companies”. The question is whether it was enough that the companies were associated companies immediately before they each ceased to be members of the old group; or whether they must remain associated companies (within a new group) immediately after they have each ceased to be members of the old group. To relate the point to the facts in the present appeal: (i) at the time of the acquisition of the Berhad shares on 16 March 1978, DIAG and Moorgate were associated companies, (ii) DIAG and Moorgate continued to be associated companies until each ceased to be a member of the International Group on 31 May 1978, and (iii) thereafter, DIAG and Moorgate were not associated companies. If section 278(2) required that DIAG and Moorgate remained associated companies after each ceased to be a member of the International Group on 31 May 1978, then it did not remove the present case from section 278(1) and (3).

The judge held that it was necessary, in order to invoke section 278(2) ICTA 1970, that the companies remained associated companies immediately after each ceased to be a member of the old group. He said this, at page 473h-474d in the report of his judgment:

Reading the definition of ‘associated companies’ in s 278(4)(a) into s 278(2), the subsection reads:

‘Where two or more companies who by themselves would form a group of companies cease to be members of the group at the same time . . .’

The question arises as to what is the relevant time at which ‘they would form a group of companies’ and in what eventuality. The language is capable of supporting both suggested constructions. I prefer, however, the construction adopted by the Revenue for two reasons. (1) The language of the section requires focus to be made on the moment when the companies cease to be members. That is the moment when the charge created by s 278(3) arises, and that is surely the moment when the conditions in the relieving provision have to be satisfied. It is at that time that the companies leaving the group must be associated, i.e. by themselves ‘would’, outside the group they are leaving, form a group of companies. The word is ‘would’: there are not the words ‘do or would’. The word ‘would’ affords some indication that the draftsman had in mind that the group that ‘would be formed’ will be different from the group that is ceasing to exist. (2) The purpose behind s 278(2) accords with this construction. The purpose must plainly be that the benefit of deferral of the tax charge conferred by s 273 could safely and sensibly be continued in respect of an acquisition by one group company from another, notwithstanding their cesser of membership of that group, if at the same time they would form in whole or in part a new group. As the commissioners put it in para 23 of their decision -

‘. . . if an asset is transferred between a sub-group within a principal group, the deferral can continue if that sub-group is sold out of the principal group. If it is appropriate to defer the tax implication of gain or loss arising on the transfer of assets

within the group, then it seems equally appropriate to allow that deferral to continue for so long as the asset remains within its own sub-group’

This, however, requires that on the cesser of membership of the previous group one of the companies shall be the 75% subsidiary of the other and both shall continue resident in the United Kingdom."

Mr Milne, of course, submits that if it is right to answer the question “was the company a company when it ceased to be a member of the group” by reference to the circumstances immediately before it ceased to be a member - as I have held in the context of examining whether condition (ii) is satisfied - then the same test should be applied to the question “were the companies associated companies when each ceased to be a member of the group”. Applying that test, the answer to the question is affirmative; and that is sufficient to invoke section 278(2) and to exclude section 278(3). It is unnecessary, and impermissible, to consider the circumstances immediately after each of the formerly associated companies ceased to be a member of the old group. The Revenue cannot be allowed to apply one test to section 278(3) and a different test to section 278(2).

There is a superficial attraction in Mr Milne’s argument; not least in its apparent simplicity and consistency. But I am satisfied that it cannot be sustained. It is, I think, necessary to keep in mind the mischief which section 278 ICTA 1970 was enacted to counter; and to ask what role in that scheme the relieving provision in section 278(2) could have been intended to have. Section 273(1) provides for the deferral of tax on gains arising on a transfer between companies within the same group. Section 278(3) brings to an end that deferral when the company which owns the asset (T in the example given by Mr Justice Millett in NAP Holdings UK Ltd v Whittles ) leaves the group. The object is to prevent the transferee company from taking the asset out of the group in circumstances in which the gain will not crystallise on a subsequent disposal - because there will be no subsequent disposal. That object is achieved when T and P (the transferor company and, in Mr Justice Millett’s example, the principal company of the group of which P and T are the only members) each cease to be members of that group on the sale of T to a third party purchaser - provided always that section 278(3) applies in that case . But, if Mr Milne’s construction is correct, that is the obvious case in which section 278(2) will exclude section 278(3). P and T would be associated companies before the sale of T to the third party purchaser; and the fact that they would not be associated companies after that sale is, as Mr Milne contends, irrelevant. It is, of course, possible that an anti-avoidance provision may miss its target; but it is difficult to believe that Parliament could have intended the relieving provisions in section 278(2) to have the effect that section 278(3) did not apply to the most obvious target at which they were aimed - the simple two company group where the asset had been transferred by the principal P to the subsidiary T in order to take advantage of the benefits conferred by section 273(1) ICTA 1970 on a sale of T to a third party. It is, therefore, both necessary and appropriate to ask whether there is some equally tenable construction of section 278(2) which does not have that effect; but which does give effect to some intention which may sensibly be attributed to the legislature. The Revenue’s construction satisfies that test. It applies section 278(2) to a case where P and T are members of a sub-group within a larger group of which, say H, is the principal company. On a sale by H of the sub-group comprising P and T the conditions of section 278(2) will be satisfied. In such a case P and T are associated companies both before and after they each cease to be members of the old group of which H is, or was, the principal company. The latent gain on the transfer of the asset by P to T is preserved within the new group of which P and T are members. The gain will crystallise in the future if T sells the asset, or if P sells T, to a third party purchaser. The object of section 278(3) is preserved.

In my view the judge was correct, for the second of the reasons which he gave, to hold that condition (iii) is satisfied.

I would dismiss this appeal.

LORD JUSTICE PILL: I agree.

LORD JUSTICE PETER GIBSON: I also agree.

Order: Appeal dismissed with costs; case remitted back to the Special Commissioners unless the figures are agreed; application for permission to appeal to the House of Lords refused. ( This order does not form part of the approved judgment )