Clive Bowring, Juliet Bowring (A vulnerable adult, acting through Clive Bowring) v The Commissioners for HM Revenue and Customs: [2015] UKUT 0550 (TCC) [2015] UKUT 0550 (TCC)

UPPER TRIBUNAL
TAX AND CHANCERY CHAMBER
[2015] UKUT 0550 (TCC)Case No FTC/115/2013
Between
Clive Bowring, Juliet Bowring (A vulnerable adult, acting through Clive Bowring)AppellantThe Commissioners for HM Revenue and CustomsRespondent
Before
Mr Justice BarlingDate 12 October 2015Category: Tax
[1]This is an appeal by Clive Bowring (“CB”) and his sister Juliet Bowring (“JB”) (together “the appellants”) against the decision of the First-tier Tribunal (“the FTT”) on 25th June 2013. In that decision the FTT dismissed the appellants’ appeals against closure notices by HMRC dated 16 May 2007 containing amendments to the appellants’ self-assessment tax returns for 2002-03. The amendments were to the effect that the appellants are liable to Capital Gains Tax (“CGT”) of £849,644 and £317,417 respectively as a result of additional gains under the Taxation of Chargeable Gains Act 1992 (“TCGA”) s.87 and supplemental charges under TCGA s.91.[2]Neither the appellants nor HMRC have called into question the primary findings of fact made by the FTT. The issues in the appeal relate to what the appellants, and also to some extent HMRC, contend to be the incorrect application by the FTT of the relevant legislation to those findings of primary fact. Whilst the appellants contend that the errors of law made by the FTT mean that the appeals must be allowed, HMRC seek to uphold the outcome determined by the FTT, albeit on different grounds from those relied upon by the FTT.[3]Before providing an outline of the facts, it is convenient to set out at this stage some of the main provisions of the legislation on the basis of which the parties have deployed their respective arguments before me. Relevant legislation[4]So far as relevant, sections 87-97 TCGA 1992 provide: s.87 (Attribution of gains to beneficiaries) 3(1) This section applies to a settlement for any year of assessment during which the trustees are at no time resident or ordinarily resident in the United Kingdom.(2) There shall be computed in respect of every year of assessment for which this section applies the amount on which the trustees would have been chargeable to tax under section 2(2) if they had been resident or ordinarily resident in the UK in the year; and that amount, together with the corresponding amount in respect of any earlier such year so far as not already treated under subsection (4) below...as chargeable gains accruing to beneficiaries under the settlement, is in this section and section...90 referred to as the trust gains for the year.(3) ...(4) Subject to the following provisions of this section, the trust gains for a year of assessment shall be treated as chargeable gains accruing in that year to beneficiaries of the settlement who receive capital payments from the trustees in that year or have received such payments in any earlier year.(5) The attribution of chargeable gains to beneficiaries under subsection (4) above shall be made in proportion to, but shall not exceed, the amounts of the capital payments received by them.(6) A capital payment shall be left out of account for the purposes of subsections (4) and (5) above to the extent that chargeable gains have by reason of the payment been treated as accruing to the recipient in an earlier year. S.89 (Migrant settlements etc) (1)…. (2) Where- (a) a non-resident period [defined in subs (1) as a period of one or more years of assessment for which section 87 applies to a settlement] is succeeded by a resident period [defined as a period of one or more years of assessment for each of which section 87 does not apply to the settlement], and (b) The trust gains for the last year of the non-resident period are not (or not wholly) treated as chargeable gains accruing in that year to beneficiaries, then, subject to subsection (3) below, those trust gains (or the outstanding part of them) shall be treated as chargeable gains accruing in the first year of the resident period to beneficiaries of the settlement who receive capital payments from the trustees in that year, and so on for the second and subsequent years until the amount treated as accruing to beneficiaries is equal to the amount of the trust gains for the last year of the non-resident period. (3) Subsections (5)…of section 87 shall apply in relation to subsection (2) above as they apply in relation to subsection (4) of that section. S.90 (Tranfers between settlements) 4 (1) If in a year of assessment for which section 87 applies to a settlement (“the transferor settlement”) the trustees transfer all or part of the settled property to the trustees of another settlement (“the transferee settlement”) then, subject to the following provisions- (a) if section 87 applies to the transferee settlement for the year, its trust gains for the year shall be treated as increased by an amount equal to the outstanding trust gains for the year of the transferor settlement or, where part only of the transferor settlement is transferred, to a proportionate part of those gains; (b) if subsection (2) of section 89 applies to the transferee settlement for the year (otherwise than by virtue of paragraph (c) below), the trust gains referred to in that subsection shall be treated as increased by the amount mentioned in paragraph (a) above; (c) if (apart from this paragraph) neither section 87 nor section 89(2) applies to the transferee settlement for the year, subsection (2) of section 89 shall apply to it as if the year were the first year of a resident period succeeding a non-resident period and the trust gains referred to in that subsection were equal to the amount mentioned in paragraph (a) above. (2) Subject to …., the reference in subsection (1)(a) above to the outstanding trust gains of the settlement is a reference to the amount of its trust gains for the year so far as they are not treated under section 87(4) as chargeable gains accruing to beneficiaries in that year. (3) … (4) This section shall not apply to a transfer so far as it is made for consideration in money or money’s worth. (5) This section shall not apply- (a) to a transfer to the extent that it is in accordance with schedule 4B treated as linked with trustee borrowing; or (b) to any chargeable gains arising by virtue of that schedule. S.97 (Supplementary provisions) (1) In sections 86A to 96 and Schedule 4C and this section “capital payment” – (a) means any payment which is not chargeable to income tax on the recipient…… (2) In subsection (1) above references to a payment include references to the transfer of an asset and the conferring of any other benefit, and to any occasion on which settled property becomes property to which s.60 applies. (3) …. 5 (4) For the purposes of sections 86A to 96 and Schedule 4C the amount of a capital payment made by way of loan, and of any other capital payment which is not an outright payment of money, shall be taken to be equal to the value of the benefit conferred by it. (5) For the purposes of sections 86A to 90 and Schedule 4C a capital payment shall be regarded as received by a beneficiary from the trustees of a settlement if- (a) he receives it from them directly or indirectly, or (b) it is directly or indirectly applied by them in payment of any debt of his or is otherwise paid or applied for his benefit, or (c) it is received by a third person at the beneficiary’s direction.[5]It is common ground that ss.90(5) was added by the Finance Act 2000 with a view to neutralising a tax avoidance scheme which is referred to by the FTT in their judgment as “Flip Flop Mark 1”, that because of the way in which ss.90(5)(a) was drafted it gave rise to another avoidance scheme known as Flip Flop Mark II, and that when HMRC became aware of this unintended effect ss.90(5) was replaced by the Finance Act 2003, too late to have any effect on the issues in this appeal. The essence of Flip Flop Marks I and II are helpfully explained in the judgment of the FTT. I gratefully quote the FTT’s explanation.[6]The FTT begin by pointing out that the effect of s.87 was to attribute to resident beneficiaries of non-resident trusts certain gains which arose within the trust to the extent that the beneficiary received capital payments from the trustees. The FTT then set out ss.90(1) and (2) and continue: 4…..This section [ie s.90] ensured that s87 continued to apply where the settlement transferred the assets to another non-resident trust and that second trust made capital payments, by providing that the transferee trust was treated as having added to it the undistributed gains of the transferor settlement. But it also ensured that section 87 could not be avoided by a transfer to a UK resident settlement. The UK settlement was treated as if it were an immigrant settlement within section 89 so that capital payments made by it were matched with the undistributed gains of the transferor settlement.[5]A planning scheme had evolved to circumvent the settlor-interested trust rules in ss77 (and 86) TCGA. These rules provided that gains of a settlorinterested trust were taxed at (normally) 40%. Colloquially it was known as the flip-flop scheme although we were not told the reason why. In brief, the scheme was that a settlor-interested trust holding assets which had substantial but as yet unrealised gains (as the assets had not been sold) would use the assets as security to borrow funds. It would then transfer the borrowed funds to a second 6 trust to make distributions to the [sic], and exclude the settlor from any benefit. In the following tax year the assets, and the gains, would be realised by the first trust, and the loans repaid out of the proceeds. At that point the first trust was only liable at 25%.1[6]The Government legislated in the Finance Act 2000 to block flip-flop planning. It introduced Schedules 4B and 4C to the TCGA. The effect of Sch 4B was to create a deemed disposal of the assets of a trust where a transfer made by a trust was linked with borrowings. This prevented the flip-flop scheme being effective as it meant that the latent gains were crystallised on the borrowing against the assets so that the gains could still be taxed on the settlor or (in a s 87 case) on beneficiaries. The effect of Sch 4C was, in a s 87 case to which Sch 4B applied, to match the Sch 4B gains with distributions to the beneficiaries in place of s 87. Flip-flop mark II[7]Ironically, however, the package of anti-avoidance provisions gave rise to a new version of the flip-flop scheme. This was because, as part of the antiavoidance measures, s90 was amended by the addition of subsection (5) so that it would not apply to transfers of value linked with trustee borrowings: ..... The reasoning, presumably, was that this kept the gains locked in the first settlement which could then be attributed to the beneficiaries under the new schedules.[8]However, the unintended consequence of s 90(5) was that it meant that nonresident trustees could choose to deliberately “switch off” the s 90(1) transfer of gains from one trust to another trust by deliberately triggering the disapplication provisions of s 90(5). This would not be advantageous if the gains remained latent at the time of distributions to the beneficiaries, as this would be caught by the new anti-avoidance provisions.2[9]The new trick required that the trust be non-resident and that there were actual gains which had not been distributed but no latent gains. For the scheme to work, the funds would also be transferred to a new trust from which they would be distributed, but before the transfer took place s 90(1) would be switched off by entering into the exact steps which would trigger the antiavoidance legislation that is Sch 4B, linking the transfer to the new trust with trustee borrowings. The intention was that the realised gains would be left behind in the original trust and the new trust could make distributions to the beneficiaries free of liability to CGT. Although Sch 4B applied in theory, it bit on nothing as there were no chargeable assets in the old trust. This planning was known as flip-flop mark II. 1 Section 90 did not apply to the realised gains as the transfer had occurred in an earlier year. 2 However, the provision in question (ss.90(5)(a) ) was drafted (presumably unintentionally) so as to disapply s.90 even where there were no significant latent gains to be triggered by schedule 4B, the gains having already been realised at the time of the transfer. 7[10]And HMRC accept that the effect of executing these steps was to leave behind the gains in the first trust and therefore, to this limited extent, HMRC accept that the flip-flop mark II planning scheme worked.[11]It is not in dispute that a flip-flop mark II scheme was implemented in this case. What HMRC do not accept is that the gains left in the original settlement cannot therefore be taxed on the beneficiaries when the new trust made distributions to them….. (The footnotes in the quotation above are mine and not the FTT’s.)[7]Therefore the parties have proceeded before the FTT and before this tribunal on the agreed basis that the Flip Flop Mark II scheme here was effective to the extent that the gains realised within the 1969 trust were not transferred to the 2002 trust pursuant to s.90, but remained in the 1969 trust.[8]I will return to the decision of the FTT in due course. First it is appropriate to provide an account of the facts. The facts[9]The following summary of the salient facts is taken largely from the skeleton arguments prepared by Mr Prosser QC, who appears for the appellants and Mr Vallat, who appears for HMRC, respectively. It is intended to reflect the findings of primary fact of the FTT which, as I have said, are not challenged in this appeal.[10]In 1969, the appellants’ father established an offshore discretionary settlement for the benefit of the appellants and other members of the family (“the 1969 trust”). The sole trustee was a Guernsey resident company called Butterfield Trust (Guernsey) Limited (“Butterfield Trust”). Butterfield Trust saw the appellants as the principal persons intended by the settlor to benefit from the 1969 trust, and considered making distributions only when requested to do so by CB.[11]By the tax year 2001-02 the 1969 trust had “trust gains” within the meaning of ss.87(2) of about £3m. By virtue of s.87(4) these would be treated as chargeable gains accruing to beneficiaries of the 1969 trust who received distributions from the trustees, and would give rise to CGT charges at a total rate of up to 64%.[12]In February 2002, CB’s solicitor, Mr Iain Whiteford of Clifford Chance, prepared a written memorandum entitled “CGT Avoidance Proposal”. The memorandum noted 8 that it was likely that the trustees of the 1969 trust would make further distributions to the appellants, possibly up to the whole of the value of the trust fund, and suggested that a Flip Flop Mark II scheme should be implemented before such distributions were made.[13]The proposed scheme and its intended effect would be as follows: the trustees of the 1969 trust would borrow money on the security of the trust fund, and then transfer the borrowed money to a second settlement with the same beneficiaries. In those circumstances the transfer to the second settlement would be treated by schedule 4B TCGA as a “transfer of value…. linked with trustee borrowing”, so that ss.90(5)(a) would prevent section 90 from applying to the transfer. Accordingly the second settlement would not “inherit” any of the trust gains of the 1969 trust pursuant to ss.90(1), with the result that the £3m trust gains would remain in the 1969 trust, allowing distributions to be made by the trustees of the second settlement to the beneficiaries of that settlement free of CGT.[14]Mr Whiteford gave a copy of his memorandum to Butterfield Trust, which agreed to go along with the scheme. In March 2002, Butterfield Trust encashed the trust fund (in part, by selling some assets to the appellants, who funded the purchases with loans from another trust) and used the proceeds to acquire gilts. By a deed dated 27 March 2002 CB3 created a new settlement, with the same beneficiaries and with Mr Whiteford and himself as the trustees, with a view to its receiving a transfer of funds from the 1969 trust (“the 2002 trust”). The terms of the 1969 trust and the 2002 trust were materially the same, and comprised part of the same settlement for the purposes of the rule against perpetuities.[15]On 2nd April 2002, Butterfield Trust borrowed £3.8m on the security of the gilts and paid that £3.8m to the trustees of the 2002 trust.[16]At the time of the transfer of the funds to the 2002 trust, Mr Whiteford expected that the whole or substantially the whole of the trust fund would be distributed to the beneficiaries shortly afterwards. However, the FTT found that “that was an expectation only: there was no definite plan of which he was aware for specific distributions to be made to beneficiaries”. Once the scheme was proposed “there was 3 See footnote 4 below. 9 no need… to discuss any plans on specific distributions: it was understood that the flip flop was there to facilitate future distributions that were likely to occur but had not yet been decided upon”.[17]Butterfield Trust adopted Mr Whiteford’s plan in the expectation that substantial distributions out of the 2002 trust, quite possibly to the entire value of the trust fund, would be made in the relatively near future. However, the FTT found that “there was no evidence whatsoever to suggest that the transfer was made on condition that certain future distributions would be made”; and “the Butterfield Trust had no say in what the trustees of the [2002 trust] chose to do”. Further, “[t]here was no agreement or understanding between Butterfield Trust and the trustees of the [2002 trust] as to what distributions would be made.”[18]According to the FTT it was CB’s intention and expectation from the outset that the funds in the 2002 trust would be distributed in the relatively near future, principally to the appellants, including in particular to enable the appellants to repay their loans, but the FTT stated that “no final decision had been taken on exactly what distributions would be made and on what dates”. Subsequently, distributions “were made on the decision of the trustees of the [2002 trust] alone”. “The final decisions on the dates and amounts of distributions were not taken by the trustees of the [2002 trust] until after [they] had received the trust funds”.[19]The following distributions, totalling £2.4m, were made by the trustees of the 2002 trust :(i) on 28th April 2002, £1,260,935 to CB and £628,706 to JB;(ii) on 15th June 2002, £58,750 to CB and £13,750 to JB;(iii) on 9th December 2002, £400,000 to CB, on the understanding (but without any legal obligation) that he would pay £200,000 each to two other beneficiaries of both the trusts, his cousins Miss Seward and Miss Pope, which he did ;(iv) on 23rd February 2003, £21,000 to JB. 10[20]Thus, at the end of the tax year 2002-03, there was £1.4 million remaining in the 2002 trust, so that less was actually distributed than originally envisaged when the scheme was proposed. The proceedings before the FTT[21]Before the FTT (as before me) it was common ground between the parties that for the purposes of ss.97(5) TCGA (above) the distributions to CB and JB could not be regarded as received from the trustees of the 2002 trust and also from the trustees of the 1969 trust. However, the FTT did not agree. The main reason the parties took, and have maintained, this view is because they consider that if the same distribution were to be regarded as received from more than one settlement, there could be liability for multiple charges to CGT on that distribution, which Parliament could not have intended.[22]If, as the appellants submitted, the source of the distributions in question was the 2002 trust alone, it was (and remains) common ground that no tax liability was incurred on those distributions. This is because the £3.8m transfer by the trustees of the 1969 settlement to the trustees of the 2002 settlement was a “transfer….treated as linked with trustee borrowing” within schedule 4B TCGA, with the result that by virtue of the (unintended) effect of ss.90(5)(a), s.90 did not apply to the transfer, and therefore the £3m trust gains of the 1969 trust could not be treated by ss.90(1)(c) and ss.89(2) as chargeable gains accruing to beneficiaries of the 2002 trust who received capital payments from the trustees of that trust. In effect, the trust gains were “locked” into the 1969 trust.[23]However if, as HMRC submitted, the source of the distributions in question was the 1969 trust alone, then it was (and remains) common ground that tax liability was incurred on the distributions. This was because the distributions would in that event be matched with the historic capital gains of the 1969 trust pursuant to s.87, s.90 not being in play.[24]In view of this, each side advanced arguments to the FTT as to why in accordance with the relevant legislation the distributions to which I have referred at paragraph 19 above fell to be regarded as capital payments received from (in the appellants’ 11 submission) the trustees of the 2002 trust alone or (in HMRC’s submission) the trustees of the 1969 trust alone.[25]As far as the payment of £400,000 made to CB on the understanding that he would pay £200,000 to each of his cousins (see paragraph 19 (iii) above), before the FTT it was common ground that that capital payment could not be regarded as received by both CB and the cousins. The reason was that Parliament could not have intended to impose multiple tax charges in respect of the same distribution. The appellants argued that the distribution should be regarded as received by the cousins rather than by CB. HMRC did not argue to the contrary or concede the point. They submitted that the FTT was entitled to find in accordance with the appellants’ submission. The decision of the FTT[26]The FTT rejected both parties’ submission that the distributions in question could not be regarded as received from the trustees of the 2002 trust and from the trustees of the 1969 trust. They held that ss.97(5)(a) should be interpreted widely, so that a capital payment can be regarded as received by a beneficiary from the trustees of one settlement directly and from the trustees of another settlement indirectly. In their view ss.87(5), purposively construed, would prevent any multiple taxation which might otherwise occur.[27]The FTT’s approach to the interpretation of the relevant statutory provisions is encapsulated the following important passages of their judgment: 121…As we have said, “from…directly or indirectly” was clearly intended to be wide in meaning; s 97(5) is clearly intended as an anti-avoidance provision and intended to catch more than a transfer of funds direct from trustee to beneficiary. As the rule in Pilkington recognises, a second trust, created under a power given to the trustees of the original trust and comprising the funds of the original trust and for the benefit of the same beneficiaries, even if the terms of the trust are different, is really part of the original donation.[122]We have already commented on the breadth that Parliament intended s 97(5) to have. Parliament cannot have intended a transfer between trusts to defeat the application of the antiavoidance provisions of ss 86 to 97 and “from … indirectly” should be read in that context. While the trustees of the original settlement cannot control when new distributions are made by the trustees of the new settlement, nevertheless they have facilitated these distributions by putting the trustees of the new settlement in possession of the funds which can only (or viewed realistically will only) be used for the benefit of the beneficiaries for the original trust.[123]In a case where the funds trace back from the new settlement to the original settlement, the new settlement is created under a power given to the trustees of the original trust, and 12 realistically the funds will be used for the benefit of the same beneficiaries, even if the terms of the trust are different, the new settlement is really part of the original donation. Therefore, on a realistic view of these facts, the new settlement is essentially the same as the original settlement, and so we would see all distributions from the new settlement as being “from … indirectly” the trustees of the original trust.[124]… Timing[125]Mr Vallat for HMRC considers that it is not enough for the new settlement simply to continue the bounty of the original settlement: he thinks there must be a plan to distribute. For this reason he says that whenever the 2002 Settlement comes to distribute the remaining funds, which has to be at least 10 years after the transfer to it from the 1969 Settlement as we are in 2013 considering the position in 2003, it would not be possible to see the distribution as being from 1969 Settlement.[126]We do not agree with the logic of this. It is irrelevant on the facts of this case as the law has now changed and distributions are no longer subject to these provisions, but taking a hypothetical situation and assuming that the law today is what it was 10 years ago, we would still see any current distribution from the new settlement as being indirectly from the original settlement in any case where the new settlement essentially was the old settlement.[28]Applying their approach to the present case, the FTT held that the distributions to CB and JB by the trustees of the 2002 trust could “clearly be traced back to” the 1969 trust. They stated that the 2002 trust had been created under a power in the 1969 trust deed4; CB and JB were in practice treated as the principal beneficiaries of both settlements; the term of the 2002 trust was co-extensive with that of the 1969 trust; nothing changed with the creation of the 2002 trust other than the identity of the trustees and some immaterial changes in the terms of the discretionary trusts; and one of the trustees of the 2002 trust was CB, who was the beneficiary in accordance with whose wishes the trustee of the 1969 trust was accustomed to make distributions. The FTT therefore held that “viewed realistically” the 2002 trust was “essentially the same as” the 1969 trust, and the distributions to CB and JB by the trustees of the 2002 trust “could not and would not have taken place except for the transfer of funds by the trustee of” the 1969 trust to the 2002 trust.[29]In the alternative, the FTT held that if (as HMRC had contended but the FTT disagreed) distributions from a new settlement can only be regarded as indirectly received from the trustees of the original settlement if the trustees of the original settlement participate in a plan for such distributions to be made, then that 4 In their decision the FTT stated that the trustee of the 1969 trust created the 2002 trust. It is common ground that this was a misstatement, and that the 2002 trust was created by CB. However, neither side placed any reliance on this error. 13 requirement was satisfied in the present case. In the FTT’s view, it was enough for the plan, which must be the plan of the trustee of the original settlement, who stands in the shoes of the settlor and controls the funds, to be “to facilitate distributions.” Here the trustee of the 1969 trust had adopted and implemented Mr Whiteford’s plan, by realising assets, borrowing funds and advancing the loaned funds to a new trust, all with the plan and object of facilitating the distribution of funds held in the 1969 trust free of tax to beneficiaries of the 1969 trust.[30]As to the £400,000 distribution to CB which he thereafter paid to the two cousins, the FTT held that, for the same reasons, this too was a capital payment received by him from the trustee of the 1969 trust indirectly, and was therefore taxable in his hands. The FTT held that the cousins did not receive a capital payment directly or indirectly from the trustees of either settlement. Although the source of the money could be traced, CB was under no legal obligation to pay it over to the cousins, and when it was received by CB it was no longer imprinted with a trust. They held that in these circumstances “causation was broken”. The issues[31]There are two issues in this appeal:(i) Whether CB and JB received the capital payments in question from the trustee of the 1969 trust within the meaning and for the purposes of ss.87(4) and ss.97(5)(a) TCGA (“Issue 1”). It is common ground that if they did, then the payments are subject to tax. (Neither in this tribunal nor in the FTT have HMRC sought to rely upon ss.97(5)(b) or (c), and the parties agree that neither of those sub-paragraphs has any application to this appeal.)(ii) Whether the £400,000 received by CB and paid by him to his two cousins was a capital payment received by CB or a capital payment received by the cousins, within the meaning and for the purposes of those provisions (“Issue 2”). It is not in dispute that if the latter is the case, it is now too late for HMRC to make an assessment in respect of the cousins. This issue only arises if the appellants are wrong on Issue 1, and the payment of £400,000 was received indirectly from the trustee of the 1969 trust for the purposes of the relevant legislation. It is common ground that if (as the appellants contend) the payment of £400,000 was received from the trustees of the 14 2002 trust alone, there would be no gains in respect of which tax could be levied, no matter who was the recipient of the capital payment. Issue 1 Summary of submissions[32]The appellants submit that the reasoning and conclusions of the FTT are wrong in law in a number of respects.[33]The appellants argue that a capital payment cannot be regarded as received from the trustees of a transferor settlement merely because the payment was made out of funds previously held by the trustees of that settlement and the beneficiaries and terms of the transferee settlement are the same as those of the transferor settlement. HMRC agree that the FTT were in error in this regard, and submit that more than this is required, in particular there needs to be a “plan”.[34]The appellants, together with HMRC, also submit that the FTT were wrong to treat the transferor settlement and the transferee settlement as one and the same for the purposes of this legislation, whatever the position may be as far as perpetuities rules are concerned.[35]Further, the appellants submit, again with the agreement of HMRC, that contrary to the view of the FTT a particular capital payment cannot be regarded as received from the trustees of more than one settlement - from one directly and from the other indirectly. Both sides therefore submit that in a situation such as the present it is necessary to determine which of the two settlements the payment in question was “receive[d]… from” within the meaning of ss.87(4). They both also accept that a capital payment being “received from” the trustees of a settlement is simply the other side of the same coin as a capital payment being “made by” the trustees of a settlement within, for example, s.91. In other words a capital payment is received from trustees if it is made by them, and vice versa.[36]The parties agree that the identification of the settlement from which the capital payment is received is to be determined on the basis of a realistic view of the facts. 15 However, they differ as to other aspects of the test by reference to which the source of the capital payment is to be identified, and as to the result in the present case.[37]In the appellants’ submission, in a case such as the present where the capital payments are “straightforward” outright distributions of settled property held on the trusts and subject to the powers of a particular settlement (the 2002 trust) to beneficiaries of that settlement, the distributions are clearly capital payments “made by” and “received from” the trustees of that settlement, and not the 1969 trust which is a different, historical, settlement.[38]HMRC, on the other hand, submit that the capital payments were not “received from” the trustees of the 2002 trust in the sense of the legislation, but were received indirectly from the trustee of the 1969 trust, the 2002 trust being a mere conduit. This is, in particular, because the payments were made pursuant to a plan to use funds in the 1969 trust to make distributions to beneficiaries of the 1969 trust. The existence of a plan was one of three factors or “signposts” referred to by the Special Commissioner, Sir Stephen Oliver QC, in Herman [2007] STC (SCD) 571. HMRC also rely upon the two other “signposts” referred to in that case.[39]The Herman case also involved a Flip Flop Mark II scheme. There was a transfer of trust assets between non-resident and resident settlements followed by distributions of the totality of those assets to beneficiaries absolutely and the winding up of the second settlement. The Special Commissioner was there determining whether the distribution to the beneficiaries was received indirectly from the original settlement for the purposes of ss.97(5)(a). As well as the existence of a plan, the Special Commissioner had regard to whether the resident settlement served merely as a vehicle or conduit to “receive and continue the act of bounty effected by the trustees of the [non-resident settlement]”, and to how the scheme was in fact implemented. Having considered the facts in the light of these “signposts”, he concluded that the payments in question were received indirectly from the trustees of the original settlement for the purpose of ss.97(5)(a). Multiple sources possible?[40]It is to be noted that in Herman the Special Commissioner was not determining from which of the two settlements the capital payment was received, but whether it 16 was received from the non-resident one. Like the FTT in the present case, he considered that it could be received from both of them, and therefore he did not need to choose between them. Although in Herman the taxpayers pointed to what they submitted were possibilities for anomalies and double taxation should multiple sources of a payment be permissible under the legislation, HMRC (then represented by different counsel) appear to have taken a more nuanced stance than in the present case, and do not seem to have argued against the possibility of multiple sources, as they have before the FTT and before me. Further, in the report of the Special Commissioner’s decision there is no indication that his very tentative view, that ss.87(5) might provide the solution to possible double taxation,5 had been the subject of substantive submissions, still less that both parties had opposed that view, as they have here.[41]In view of the potential significance of the issue whether more than one source of payment is possible for these purposes, it is appropriate to consider it at this stage.[42]As I have said, the parties both submit that the FTT were in error in two related respects: first, in holding that a particular distribution could be “from” more than one settlement for the purposes of the legislation; and second, in holding that ss.87(5) would be capable of operating across two or more settlements, so as to remove the risk of multiple taxation which the parties agree could arise if more than one source of a capital payment were possible. It is also common ground that the FTT’s approach would be problematic for the operation of the legislation in other respects. The parties submit that the absence of a solution to the issue of double taxation is a strong indication that ss.97(5)(a) should not be construed as encompassing multiple sources of a payment.[43]As far as the possibility of double taxation is concerned, the appellants put forward the following example: the trustees of one settlement (“S1”) transfer part of the settled property to trustees of another settlement (“S2”), and subsequently the latter make a distribution to a beneficiary, B. If the distribution is regarded as received from the trustees of both settlements, then B could be subject to CGT in respect of the trust gains of S2 and of S1, giving rise to CGT at a rate of up to 128%. 5 “…it is not beyond the realms of purposive construction that s 87(5) might apply to prevent the double counting problems.” (Paragraph 20 of the Special Commissioner’s decision.) 17[44]Although such double taxation did not arise on the facts of the present case, the FTT did not suggest that it could not do so on other facts. Their solution was that the risk is prevented by ss.87(5), which provides that “the attribution of chargeable gains to beneficiaries under [ss. 87(4)] shall be made in proportion to, but shall not exceed, the amounts of the capital payments received by them.”[45]This solution would require one to construe ss.87(5) as preventing the capital gains of two or more settlements from being attributed to a beneficiary in respect of a single payment to that beneficiary. One formidable problem with this is that ss.87(1) provides that the section applies to “a settlement”. Ss.87(2) then defines the trust gains of that settlement, and ss.87(4) goes on to provide that those trust gains are treated as chargeable gains accruing to beneficiaries of “the settlement” who receive capital payments from the trustees of that settlement. I agree with both the parties that these provisions are clearly concerned only with the attribution of the trust gains of a particular settlement. In these circumstances it is difficult to see how ss.87(5), which applies expressly to the attribution of gains “under subsection (4) above” ie to the attribution of gains in respect of the same settlement, could be construed as applying across two or more settlements. This problem of construction can only be exacerbated where one is a settlement to which s.87 applies, and the other is a settlement to which ss.89(2) applies.[46]The parties pointed to other significant difficulties with the FTT’s reliance on ss. 87(5) as a solution to the double taxation risk. One such difficulty relates to the year of assessment. By ss.87(1), “This section applies to a settlement for any year of assessment…”. Similarly, ss.87(2) to (4) are concerned only with the trust gains of the settlement for that year. However, using the appellants’ example at paragraph 43 above, a beneficiary could be taxable in respect of the trust gains of S1 in one year and the trust gains of S2 in a different year. I agree that it is difficult to see how ss.87(5) could be applied to different settlements so as to prevent the attribution of the gains of those settlements in respect of a single capital payment, given that each settlement accrued its gains in a different year.[47]Further, both parties submit that if ss.87(5) can apply to the attribution of the trust gains of more than one settlement in the context of a single capital payment, then one needs to identify the settlement(s) whose gains are to be attributed to the beneficiary, 18 and if both then to what extent. Identification of the settlement is important, not least for the purpose of applying the increased rate of tax imposed by sections 91-95 (which increase the tax charge the longer the delay between the accruing of the gains to the settlement and the making of the capital payment). It is also necessary for the purpose of deciding how much, if any, of the gains of a particular settlement are still available to be attributed to beneficiaries in receipt of other capital payments under s.87 or ss.89(2), or are available to be transferred to a transferee settlement under s.90. The parties point to the fact that ss.87(5) provides no means for such identification to be made. The appellants, in addition, refer to a number of other questions which would be raised. For example, is the attribution of gains to be made on a pro rata basis depending on the extent of each settlement’s gains? Does a direct receipt from one settlement take priority over an indirect receipt from another settlement? If the trust gains of the settlements are for different years, should the earlier gains have priority?[48]It is clear in my view that the legislation does not provide answers to any of these questions, and that the courts would be legislating if they were to attempt to do so. The parties submit that ss.87(5) is silent on these matters because they do not arise, as the provision cannot properly be construed so as to operate across more than one settlement in respect of a single payment to a beneficiary.[49]The FTT agreed that the legislation contained provisions intended to prevent double taxation. They noted that ss.87(6) was intended to avoid the double taxation that might otherwise arise under ss.87(4) if it was applied to the same beneficiary in respect of the same gains in different years, and that ss 87(6) would be unnecessary if ss.87(5) had been intended to apply across different tax years. The FTT therefore thought that there was “something in” the parties’ submissions on the interpretation of ss.87(5). However, they nevertheless agreed with the Special Commissioner in Herman that, read purposively in the context of sections 86-97, which showed that Parliament intended that beneficiaries should only be taxed to the extent that distributions were actually received, ss.87(5) should be construed as requiring that 19 any attributions to any person under ss.87(4) cannot exceed the amounts of capital payments received by them from any trust in any year.6[50]I have already noted that the Special Commissioner’s view on this point, on which the FTT placed reliance, was expressed in the most tentative terms, nor is there any indication that he had the benefit of the detailed submissions which have been put to me.7 In my view, for the reasons set out above it is not possible to construe ss.87(5) as operating other than in respect of receipts from a single settlement in a single year. I do not therefore consider that it can provide a solution to the risk of double taxation which could admittedly arise (albeit it did not in this case) if a single capital payment can be “received from” more than one settlement for these purposes. It follows that the FTT erred in law in that regard.[51]No other solution to this risk has been identified. This means that despite the intention of Parliament, clearly expressed through ss.87(5) and (6), double taxation remains possible if, as the FTT held, ss.97(5)(a) permits the same capital receipt to be treated as having been received both directly from the trustees of one settlement and indirectly from the trustees of another settlement.[52]Like the parties, I do not consider that ss.97(5)(a) can be construed in that way. The absence of any means of avoiding the risk of multiple taxation, and the other anomalies and uncertainties to which that construction would give rise, is a powerful indication against the adoption of it.[53]Further, the provision itself, together with its related provisions in ss.97(5)(b) and (c), appear on their plain terms to be aimed at ensuring that a beneficiary falls to be treated as in receipt of a relevant payment where the legislation regards such treatment as appropriate, rather than expanding the sources from whom the payment is to be regarded as received. The wording of ss.97(5)(a) is also framed in terms which naturally appear to envisage that a particular payment should be “received …from” a single trust, either directly or indirectly. It does not sit comfortably with the provision for a single payment to be regarded as received from more than one settlement. 6 Paragraphs 173-181 of the FTT’s decision. 7 See paragraph 40 above and footnote 5 thereunder. 20[54]Nor, quite apart from the problem of multiple taxation which would arise, is there any obvious reason why Parliament should be taken to have intended such a result. The main thrust of the legislation of which ss.97(5)(a) forms part is to specify with precision in which circumstances trust gains of a particular settlement are to be treated as attributable to the beneficiaries of that settlement who are in receipt of distributions, and in which circumstances the trust gains of a particular settlement are to be treated as transferred to a separate settlement so that they can be matched to a payment made to a beneficiary of that second settlement. It would be inconsistent with such arrangements, and unnecessary, to treat a distribution within the scope of these provisions as being received from more than one settlement.[55]In my view the natural meaning and effect of indirect receipt in ss.97(5)(a) is that a beneficiary is still to be taken to have received a payment from a settlement even if it is paid to him through an intermediary. It is conceivable that a separate settlement could, depending on the circumstances, constitute such an intermediary, but the payment would not in those circumstances fall to be treated as “made by” and “received from” that intermediary for the purposes of the legislation.[56]For these reasons I consider that the FTT erred in holding that ss.97(5)(a) permits the same capital payment to be treated, for the purposes of the legislation, as having been “received from” the trustees of one settlement directly and also from the trustees of another settlement indirectly. Were the payments received from the 1969 trust or the 2002 trust?[57]Because of their conclusion on the previous point, the FTT did not specifically consider and determine which one of the two settlements the payments were “received from” and “made by” in this case. However, both parties invited me, if I accepted their submissions on the multiple sources construction point, to exercise the power under ss.12(4) of the Tribunals, Courts and Enforcement Act 2007 and decide this issue myself rather than remitting it to the FTT. In support of that course they emphasised that the findings of primary fact made by the FTT are not in dispute, and submitted that these findings provide a sufficient basis for me to determine the issue.[58]I agree that this is an appropriate course to take. I see no merit in remitting the matter. The FTT’s findings of primary fact are full and meticulously considered, and 21 although they did not find it necessary to make a determination as between the two settlements, the FTT’s approach to determining whether the payments were “received from” the 1969 trust is clearly set out in their decision.[59]That approach, which I have already described,8 differed from the tests put forward respectively by the appellants and HMRC.9 In essence the FTT took the view that a distribution could be regarded as “received from” the trustees of a transferor settlement if they have “facilitated” the distribution by putting the trustees of the new settlement in possession of funds which, viewed realistically, will only be used and are used for the benefit of the beneficiaries of the original settlement, who are also the beneficiaries of the transferee settlement. In those circumstances the funds “trace back” from the new to the original settlement, and the new settlement is “really part of the original donation” so that the two settlements are essentially one and the same, with the result that all distributions from the new settlement are to be regarded as being “from … indirectly” the trustees of the original trust.[60]Neither party has sought to support that approach and both have submitted that it is wrong. In particular both agree that the two settlements cannot be regarded as the same settlement for tax purposes. Both also submit (although for different reasons) that a payment which is actually made to a beneficiary by the trustee of the transferee trust does not fall to be treated as “received from” the trustee of the transferor trust merely because the latter has facilitated the distribution by putting the trustees of the transferee settlement in possession of funds which, on a realistic view of the facts, will only be used for the benefit of the beneficiaries of the transferor settlement, who are also the beneficiaries of the transferee settlement.[61]The appellants submit that, on the basis of the facts as found, it is “impossible” to regard the capital payments in question as other than made by and received from the 2002 trust. According to HMRC, on the other hand, it is necessary to determine which one of the two settlements the payment is “really” from by reference to the three “signposts” in Herman. In Mr Vallat’s submission these all point in the direction of the payments being indirectly from the 1969 trust, in particular because the funds originated from that settlement, and the distributions were made pursuant to a plan 8 See paragraphs 27 to 29 above. 9 See paragraphs 37 and 38 above. 22 participated in by the trustees of that settlement. As I have said, the FTT expressly rejected the need for a plan, but stated that if it was a requirement it was satisfied in this case (see paragraph 29 above).[62]Mr Prosser emphasised certain aspects of the FTT’s undisputed findings of fact: the trustee of the 1969 trust had parted with its assets to the 2002 trust with no conditions or agreement as to what would be done with them – there were no strings attached to the transfer of the assets, and the trustee of the 1969 trust had no say in what the trustees of the 2002 trust decided to do; the capital payments to the beneficiaries were made out of the settled property of the 2002 trust; the trustees of the 2002 trust (as well as the trustee of the 1969 trust) were participants in a plan but the 2002 trustees alone decided to make the capital payments in exercise of their own discretion, and they alone decided on the amounts and dates of the payments; those decisions were taken only after the funds had been transferred to the 2002 trust; and in the event significantly less was distributed than had originally been envisaged (£2.4 million out of a total of £3.8 million transferred).[63]In these circumstances Mr Prosser submitted that as the capital payments were outright distributions of settled property, held on the trusts and subject to the powers of the 2002 trust, made to beneficiaries of that settlement, it is impossible to regard the distributions as other than “received from” the trustees of the 2002 trust. He submitted that it is also impossible to regard the capital payments as received from the trustees of the 1969 trust, and it is irrelevant that, as a matter of history, the funds in question originated from the 1969 trust and the trustee of that trust participated in a plan.[64]Mr Prosser argued that these submissions are confirmed by the legislation itself in several respects. (i) Settlement must exist when distribution is received?[65]In a somewhat convoluted argument, he referred first to s.87, which he said presupposed that the trustees and the settlement are in existence in the year in which the capital payment is received, including where the payment is received indirectly. For this pre-supposition he relied upon several features of the section. 23[66]Ss.87(4) attributed trust gains of a settlement for a particular year being, by virtue of ss.87(1), a year in which the trustees are at no time resident in the UK. In this respect the argument appeared to be that in order to be non-UK resident the trustees/settlement must be resident elsewhere and therefore in existence.[67]Next the argument focussed on ss.87(2) and (4). Under these one is required to consider on what gains the trustees would have been taxable if they had been UKresident in the year in question. If a capital payment has been received by a beneficiary in that year or in an earlier year, then the gains are matched to that payment. Therefore, so the argument goes, if the trustees/settlement must exist in the year in which the trust gains accrue (as it is assumed must be the case), then it follows that they must have existed in an earlier year when the capital payment was received. The settlement’s existence at the time of receipt of the distribution being (in the appellants’ submission) a pre-supposition of the legislation, the possibility (on HMRC’s case) of “harking back” and attributing a capital payment to trustees of a transferor settlement like the 1969 trust would depend on whether that transferor settlement still existed at the time when the payment was received by the beneficiaries (from the transferee settlement). That point in time could be years later and the transferor’s continuing existence would therefore be fortuitous and would render the operation of the legislation wholly arbitrary. This, in the appellants’ submission, militates against a construction of ss.97(5)(a) which permits such “harking back”.[68]The FTT were not impressed by the point, and Mr Vallat submits that it is not a good one.10 In his submission there is no requirement that the settlement should exist in the year of receipt of a capital distribution, and there is no obstacle to the application of s.87 to a settlement and to trustees that no longer exist. He contends that if they no longer exist, then as a matter of language they are not resident in the UK; although there will be no further gains, the provisions, he says, work perfectly well to carry forward gains from previous years to years when the settlement has ceased to exist. 10 The FTT, at paragraph 150 of their decision, record Mr Vallat as agreeing with the appellants’ argument. Mr Vallat informed me that this was not the case. His skeleton argument for the FTT did not indicate agreement, and there therefore appears to have been a misunderstanding as to HMRC’s position on the point. 24[69]Mr Vallat has a further argument against the point. He submits that if the appellants were right, there would have been an obvious and serious lacuna in the legislation: trustees could start the process of payment in one tax year but the beneficiary might only receive the payment in a second tax year, with the trust being wound up in the meantime. He illustrated this as follows: trustees decide to wind up a settlement at the end of March and, on 5th April, post a cheque or cash to beneficiary A whilst declaring that they hold any remaining assets on trust for beneficiary B absolutely. By virtue of sections 60 and 72 of this legislation the trust would end at that point. However, the cheque is not received and cashed (or the cash is not received) by A until the following tax year. Mr Vallat submitted that such a scenario was not fanciful and could easily be arranged; if it were, then tax could be avoided by a far simpler mechanism than was actually adopted here.[70]In their decision the FTT noted that the point was not of direct relevance in this case, as in fact the 1969 trust continued to exist long after the relevant capital payments were made. Mr Prosser seeks to rely upon it only as an aid to the construction of ss.97(5)(a).[71]Mr Prosser frankly acknowledged that this argument was not a show–stopper. Ingenious as it is, I do not find that it helps in construing ss.97(5)(a), and I agree with the submissions of Mr Vallat. I find it impossible to conclude on the basis of the provisions relied upon by Mr Prosser that Parliament must be taken to have intended a settlement/trustees to be in existence when a capital distribution was received, or that the presupposition of continued existence said to be found in s.87 is established as a matter of construction. If Parliament had directed its mind to this question it is very likely that it would have wished to avoid the obvious lacuna to which Mr Vallat and the FTT refer. Finally, as the provision was drawn to my attention, I should say that I do not consider either side to be significantly assisted by ss.94(4) (which simply indicates that the draftsman of the legislation was alive to the possibility of a transferee settlement not existing in a particular year). (ii) Making/receiving capital payment are two sides of same coin 25[72]Next, Mr Prosser referred to his submission, accepted by Mr Vallat,11 that the statutory concept of “a beneficiary receiving a capital payment from trustees of a settlement” and the statutory concept of “the trustees of a settlement making a capital payment” are simply two sides of the same coin. Mr Prosser stated that s.87 uses the language of receipt because there the focus is on the beneficiary who is liable for the tax, rather than the trustee, whereas in other parts of the same statutory provisions (for example, ss.91(1), 92(3) and 93(1)) the draftsman refers to a capital payment being “made by” a trustee. He said it was obvious that the trustees who made the capital distributions to CB and JB were not the trustees of the 1969 trust but the trustees of the 2002 trust. He pointed to the FTT’s language, which described the trustees of the 1969 trust as merely “facilitating” the distributions, and recorded that “The trustees of the 2002 Settlement made distributions totalling £2.4 million…”.12 He also referred to ss.97(2), and asked rhetorically how it could be said that the payment was made by the 1969 trust if, instead of paying cash, the trustees of the 2002 trust had bought an asset with the trust funds and transferred that asset to the beneficiaries.[73]Mr Vallat accepts the “two sides of the same coin” analysis and, therefore, that for his case to succeed it is necessary that the 1969 trust “made” the capital payments to CB and JB. However, he submits that this cannot rule out indirect payments such as, on the appellants’ alternative case, the payment to the cousins via CB. In his submission all the payments in the present case were “made” by the 1969 trust for the relevant statutory purposes, and the references in the FTT’s findings of fact13 to the “making” of the distributions by the 2002 trust should not be treated as referring to the statutory purpose in issue, not least because the FTT also described the plan as involving distributions “from” the 1969 settlement.14 Ss.97(2) did not, in his view, assist the appellants, because all that provision does is to widen the concept of “payment”, so that had the trustees of the 2002 trust bought an asset and transferred it to the beneficiaries, that would still have constituted an indirect payment made by the 1969 trust.[74]Although the features of the legislation relied upon by Mr Prosser under this heading are not conclusive, the legislation’s use interchangeably of the “made by” and 11 See paragraph 35 above. 12 See paragraphs 21, 122, 137 and 139 of the FTT’s decision. 13 At paragraph 21 of their decision. 14 At paragraph 139 thereof. 26 “received from” concepts does, in my view, provide some support for the appellants’ case. So does ss.97(2): in the situation envisaged by that provision, it would require considerable straining of the concept of an indirect payment in order to regard the transfer to beneficiaries by the 2002 trustees of an asset which they had bought with their settled property, as being indirectly “received from” the 1969 trust, rather than from the 2002 trust directly. Similar points can be made in relation to ss.97(5)(b) and (c). (iii) The rationale of s.90[75]Mr Prosser’s third supporting argument relates to the rationale of s.90. He submits that the essential purpose of s.9015 is to transfer trust gains of the transferor settlement for the relevant year to the transferee settlement because the capital distribution will be treated as “received from” the trustees of the latter settlement, and not from the trustees of the transferor. The thinking behind this transfer of gains is, he says, obvious. If the trustees of the transferee settlement make a distribution to a beneficiary in a later year, that will be a capital payment “received from” (“made by”) those trustees, and it will not be a payment “received from” (“made by”) the trustees of the transferor settlement, who made the trust gains. Therefore, absent s.90, the payment made by the transferee settlement would not be matched with the trust gains of the transferor settlement, and tax would be avoided. To prevent this loss, s.90 transfers the gains to the settlement from which capital payments will be received in the future.[76]Mr Prosser emphasises that s.90 therefore recognises the separate existence of each of the two settlements, and that if the trust gains remain in the first settlement then tax will be avoided. Hence the statutory transfer. He submits that HMRC’s approach in the present case turns this rationale on its head by asserting that the capital payments were “received from” (“made by”) the trustees of the transferor settlement, when they were obviously “received from” (“made by”) the trustees of the transferee. If s.90 had had its intended effect of transferring the relevant gains to the 2002 trust, then HMRC’s current assertion of an indirect payment by the transferor would mean that the trust gains were in the wrong settlement, unable to be matched by the capital payments made to beneficiaries by the transferee settlement. The only 15 Here, the relevant provision is ss.90(1)(c) because the 2002 trust is a new UK settlement. 27 reason HMRC are running this argument is, he says, because of the flaw in the drafting of ss.90(5)(a) described earlier,16 the effect of which is that s.90 does not apply in this particular case, and therefore does not have its intended effect of transferring the trust gains.[77]Mr Prosser developed this point by submitting that if s.90 had operated as intended, and the section had transferred the realised gains of the 1969 trust to the 2002 trust, then HMRC would not be seeking to make what he termed a perverse argument, namely that the payments here were indirectly made by/received from the 1969 trust. He submitted that in a case where s.90 was not “switched off” by the unintended effect of ss.90(5)(a) and schedule 4B, for example where there was a transfer of trust assets between settlements without linked borrowing, it is inconceivable that HMRC would make the present argument, regardless of whether the transfer of trust assets was part of a plan. In those circumstances HMRC would say, with justification, that any such plan was irrelevant and that the payments were clearly received from/made by the transferee settlement. Instead, they are now seeking to strain the language of the statute unacceptably, and adopting a wholly unrealistic view of the facts, in order to avoid the obvious conclusion.[78]Mr Vallat, in response, seeks to sidestep this limb of Mr Prosser’s arguments by submitting that, on the facts of the present case, s.90 would not have applied to transfer the gains to the transferee settlement, even in the absence of the flawed drafting of ss.90(5)(a). This is because here no transfer of settled property between the two settlements has taken place within the meaning of ss.90(1). In developing this point Mr Vallat accepted that, as a matter of general law, there had been two steps:(i) a transfer of trust assets from the 1969 trust to the 2002 trust, and(ii) capital payments from the 2002 trust to the beneficiaries. But the question was whether these should be taxed as two separate steps (i.e. a transfer of settled property within s.90, and then a direct capital payment from the 2002 trust) or as a composite whole (i.e. no transfer of settled property and an indirect capital payment from the 1969 trust). If the realistic view of the facts is that one should look at the composite whole for tax purposes, the individual steps would not be taxed. 16 Paragraphs 5-7, and footnote 2 above. 28[79]In this connection Mr Vallat drew my attention to a number of formulations of the principle first set out by the House of Lords in WT Ramsay Ltd v IRC [1982] AC 300. These included: Berry v HMRC [2011] UKUT 81 (TCC), [2011] STC 1057, per Lewison J (as he then was) at paragraph [31]; IRC v McGuckian [1997] STC 908, per Lord Browne-Wilkinson at pp.913-914, and Lord Steyn at 918f; and Barclays Mercantile Business Finance Ltd v Mawson (Inspector of Taxes) [2004] UKHL 51 at paragraphs [32] and [36]. In the latter case Lord Nicholls, whose speech represented the opinion of the whole court, said:
“The essence of the new approach was to give the statutory provision a purposive construction in order to determine the nature of the transaction to which it was intended to apply and then to decide whether the actual transaction (which might involve considering the overall effect of a number of elements intended to operate together) answered to the statutory description. Of course this does not mean that the courts have to put their reasoning into the straitjacket of first construing the statute in the abstract and then looking at the facts. It might be more convenient to analyse the facts and then ask whether they satisfy the requirements of the statute. But however one approaches the matter, the question is always whether the relevant provision of statute, upon its true construction, applies to the facts as found. As Lord Nicholls of Birkenhead said in MacNiven v Westmoreland Investments Ltd [2003] 1 AC 311, 320, para 8: "The paramount question always is one of interpretation of the particular statutory provision and its application to the facts of the case." ”
[80]Having then referred to a number of other authorities Lord Nicholls went on: “Cases such as these gave rise to a view that, in the application of any taxing statute, transactions or elements of transactions which had no commercial purpose were to be disregarded. But that is going too far. It elides the two steps which are necessary in the application of any statutory provision: 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. As Ribeiro PJ said in Collector of Stamp Revenue v Arrowtown Assets Ltd [2003] HKCFA 46, para 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."[81]Mr Vallat submitted that in the light of these principles, applying the “signposts” in Herman, and taking a realistic view of the facts, the capital payments in the present case were made by the 1969 trust via an intermediary settlement, namely the 2002 trust.

(a) 29

[82]The FTT rejected this approach17, and with respect I consider that they were right to do so. In my view it cannot be sustained. It requires one to ignore the clear words (and the intended modus operandi and effect) of s.90. That provision, as part of its machinery to prevent tax avoidance, refers to a straightforward legal concept, namely a transfer of settled property. Such a “transfer” takes place if property ceases to be settled property of the transferor settlement and becomes settled property of the transferee settlement. S.90 expressly acknowledges and builds upon such a transfer by ensuring that relevant trust gains are also transferred, because (so it assumes) it is from the trustees of the transferee settlement that capital payments from the transferred property will be received. No real exercise in interpretation or construction is required in order to ascertain whether such a transfer of settled property has occurred or not. In the present case it clearly has, as HMRC accept, at least as a matter of general law. In these circumstances it is not possible, in my view, to construe s.90 in such a way that the actual transfer here is treated as not having occurred for tax purposes. Such a construction would reflect neither a realistic view of the facts nor the clear purpose and modus operandi of the legislation.[83]Further, as Mr Prosser pointed out, if such a construction depends on the application of the Herman signposts, in particular whether a plan exists, and whether and to what extent that plan is actually implemented, then various questions and uncertainties arise: How long does one have to wait to find out about implementation? If, in the meantime, the trustees of the transferor settlement are making capital payments to beneficiaries, how does one ascertain whether those beneficiaries are taxable or not, and therefore, whether and to what extent the trust gains are still in that settlement or have been transferred under s.90? What is the position here in relation to the £1.4 million which was not distributed by the trustees of the 2002 trust? Is that sum, too, to be regarded as not transferred to the 2002 trust, on the basis that the plan appears to have been to distribute all or most of the £3.8million to the beneficiaries of the 2002 trust? Or, in the case of the £1.4 million, does the failure to implement the plan to that extent mean that the transfer is to be regarded as having merely been delayed? 17 See paragraphs 183-4 of their decision. 30[84]Mr Vallat’s answer was that there was no transfer of assets to the extent that the planned distributions were given effect. This response, however, does not resolve the question of how long one needs to wait to see whether the plan has been given effect. In the present case the plan did not cover when payments would be made or their amount.[85]In my view Mr Vallat’s argument in this regard does not obtain any assistance from the Ramsay principle and the case law referred to above. As I have said, the clear purpose of s.90 is to prevent the tax avoidance which might occur where trust property is transferred from one settlement to another, followed by capital payments by the trustees of the transferee settlement to its beneficiaries, and the statutory provision achieves this purpose by transferring relevant trust gains to the transferee settlement so that they can be matched with those payments. It would therefore be wholly inconsistent with that statutory purpose and mechanism (as well as with the clear language of the legislation) to construe the statutory concept of “transfer of settled property” out of existence by reference to a plan - a feature which is not mentioned by the statute.[86]Nothing in the Ramsay case law requires or entitles one to do this. Indeed, if (to borrow the words of Ribeiro PJ in the passage cited by Lord Nicholls (above)) one asked whether s.90 and its associated provisions, construed purposively, were intended to apply to the transfer of settled property on a realistic view of the facts in the present case, the answer is clearly yes. In my view Mr Vallat’s argument also falls foul of the comment in Lewison J’s exposition of the Ramsay principle in the Berry case (above), at paragraph 31 of his judgment:
“(vi) However, the more comprehensively Parliament sets out the scope of a statutory provision or description, the less room there will be for an appeal to a purpose which is not the literal meaning of the words. (This, I think, is what Arden LJ meant in Astall v HMRC (§ 34). As Lord Hoffmann put it in an article on Tax Avoidance: “It is one thing to give a statute a purposive construction. It is another to rectify the terms of highly prescriptive legislation in order to include provisions which might have been included but are not actually there”: See Mayes v HMRC [2010] STC 1 (§ 30)).”
[87]It is clear that sections 86ff are highly prescriptive provisions, intended by Parliament to be comprehensive in their scope. Mr Vallat’s approach would go well beyond any permissible purposive interpretation and would amount to rectification of the legislation to include provisions which might have been, but were not, included. 31 HMRC’s interpretation of ss.90(1) in relation to whether or not a transfer of settled property has occurred within the meaning of that subsection, is at odds with the language, purpose and mode of operation of the provision, and would be wholly uncertain in its effects, and unworkable. Conclusion on Issue 1[88]For the reasons discussed earlier, I am free to make my own assessment of the question raised by Issue 1 on the basis of the undisputed findings of fact made by the FTT, and taking a realistic view of those facts.[89]I do not believe that assessment of the facts in a case such as this, in order to determine whether a capital distribution was received from/made by a particular settlement, is susceptible to the application of some more or less formulaic “test”, whether by reference to the existence of a plan or otherwise. In my view the issue raised here requires all relevant factors to be considered, and each case will depend on its own facts. Relevant factors will no doubt include whether what is done is pursuant to a plan or understanding or agreement.[90]In relation to the FTT’s findings, I note in particular that the trustee of the 1969 trust made an outright, unconditional transfer of its settled property to the 2002 trust. There was no agreement between the trustee of the 1969 trust and the trustees of the 2002 trust, and the former admittedly had no say whatsoever in what the latter did with the transferred property, whether by way of distributions to the beneficiaries or otherwise. It is true that there was a plan to enable capital payments to be made free of CGT, that the plan envisaged virtually all the transferred property being paid to the beneficiaries of the 2002 trust (who were also beneficiaries of the 1969 trust), and that both sets of trustees knowingly played a part in the realisation of the plan. However, the existence of the plan did not affect the fact that the 1969 trust’s settled property was transferred to the 2002 trust and was held on the terms of the latter settlement, and that when the 2002 trustees decided to make (and made) the capital payments, they did so entirely in the exercise of their own discretion - they alone decided on the amounts and dates of the payments, and they made those decisions after the trust assets had been transferred to them. This independence of decision-making is 32 underscored by the fact that they decided to leave a substantial part of the trust assets undistributed, notwithstanding that the plan had envisaged otherwise.[91]In assessing the undisputed facts, I conclude that the capital distributions in the present case were clearly received from/made by the 2002 trust, for the purposes of the legislation. I agree that Mr Prosser’s submissions on Issue 1 are very strongly supported by the legislation in the respects identified above, and in particular by the purpose and modus operandi of s.90. I do not consider it possible, taking a realistic view of the facts here, to regard the 2002 trust as a mere intermediary in the sense that would be necessary if the distributions were to be treated as received from/made by the 1969 trust indirectly.[92]This conclusion, of course, means that the aim of the legislation as a tax avoidance measure has not been achieved in the present case. However, this is because the measures, which were tailored to deal with a situation such as the present, did not have the intended effect, and admittedly failed to transfer trust gains so that they could be matched with capital payments. That being so, in my judgment it is not open to this tribunal to strain the facts, and the clear meaning and effect of the legislation, to the extent which would be necessary to produce the outcome for which HMRC argue.[93]Accordingly I determine Issue 1 in favour of the appellants. Issue 2[94]As stated earlier, this issue only arises if the appellants’ case on Issue 1 should fail. In view of my conclusion that none of the distributions in question (including the payment of £400,000 received by CB and paid by him to the two cousins) were received from the trustee of the 1969 trust, it is irrelevant whether the recipient for statutory purposes of the £400,000 capital payment was CB or the two cousins, as it is agreed that no tax charge arises. Relief[95]In the light of the above, the appeal must be allowed and the decision of the FTT is set aside. 33[96]I invite the parties to agree and submit to me for approval an order reflecting this decision, including any consequential orders or directions. The Honourable Mr Justice Barling Judge of the Upper Tribunal Release date: 12 October 2015

Cited in 1 later judgment