Bowring & Anor v Revenue & Customs [2013] UKFTT 366 (TC)

FTT-Tax

Bowring & Anor v Revenue & Customs

[2013] UKFTT 366 (TC) · 2013-06-25

[1]These appeals are against conclusions and amendments to the appellants’ self-assessment tax returns for the year 2002/03 contained in closure notices dated 16 May 2007. These closure notices amended the appellants’ returns to include additional gains under the Taxation of Chargeable Gains Act 1992 (“TCGA”) s 87 as well as supplemental charges under TCGA s 91, on the basis that the appellants had received capital payments that were to be matched with the capital gains of an offshore settlement. The effect of the amendments was to increase Mr Bowring’s liability to tax by £849,644 and Miss Bowring’s liability to tax by £317,417.68. Background – flip-flops Flip-flop Mark I

[2]Legislation has existed since 1981 which taxed UK resident beneficiaries on capital distributions received out of gains realised by non-resident trusts. The purpose of this legislation was self evidently to ensure that UK-resident beneficiaries of non-resident trusts did not have a tax advantage over beneficiaries of resident trusts: trustees of resident trusts are liable to CGT on gains and capital distributions to beneficiaries will accordingly be net of tax. As trustees of non-resident trusts are not liable to UK tax, the provisions discussed below were intended to even up the position by imposing on UK resident beneficiaries a tax charge on the non-resident trust’s gains to the extent the beneficiaries actually received distributions. The current legislation replaced an earlier regime which taxed beneficiaries on the gains of non-resident trustees as they arose irrespective of whether the beneficiaries received capital distributions. The current rules therefore make distribution the trigger to liability, and the sole issue in this case is whether there has been a distribution (within the extended sense given by the legislation) from a non-resident trust to the appellants.

[3]In the year of assessment 2002/03 S 87 TCGA provided in so far as relevant:

“Attribution of gains to beneficiaries (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 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) …. (6A) …. (7) A beneficiary shall not be charged to tax on chargeable gains treated by virtue of subsection (4) above as accruing to him in any year unless he is domiciled in the United Kingdom at some time in that year. (8) … (9) … (10) ….”

In other words, s 87 attributed to resident beneficiaries of non-resident trusts certain gains which arose in the trust only to the extent that the beneficiary received certain capital payments.

[4]Another provision, s 90 TCGA, provided: Transfers between settlements(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 settled property is transferred, to a proportionate part of those trust gains; (b) …. (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 subsection (3) below, the reference in subsection (1)(a) above to the outstanding trust gains for the year of the transferor 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. ….” This section ensured that s 87 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 ss 77 (and 86) TCGA. These rules provided that gains of a settlor-interested 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 trust to make distributions to the , 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%.

[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 anti-avoidance measures, s 90 was amended by the addition of subsection (5) so that it would not apply to transfers of value linked with trustee borrowings:

“(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.”

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 non-resident 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.

[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 anti-avoidance 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.

[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. Before we consider the law on this, we outline the particular facts in this case. The facts

[12]The appellants are brother and sister. Mr Bowring acts as attorney for his sister who is a vulnerable adult. In these circumstances, any knowledge or intentions of Mr Bowring we also treat as the knowledge and intentions of Miss Bowring. Mr and Miss Bowring have been resident, ordinarily resident and domiciled in the UK at all material times.

[13]They were beneficiaries (together with other family members) of a non-UK resident discretionary trust created in 1969 by their father and referred to in this decision notice as the 1969 Settlement. At the time relevant to this appeal, the trustee of the 1969 Settlement was Butterfield Trust (Guernsey) Limited (“Butterfield Trust”).

[14]We were not shown any letter of wishes executed by the settlor but it was accepted that the two appellants were, and were seen by the Butterfield Trust as being, the principal persons intended by the settlor to benefit from the trust. We find that, so far as making distributions to beneficiaries was concerned, Butterfield Trust was reactive: it considered making distributions only when Mr Bowring requested it to do so.

[15]By 2001/2 the 1969 Settlement had trust gains within the meaning of s 87 TCGA of about £3 million. These would be treated as chargeable gains accruing to the beneficiaries who received distributions from the trustees, which together with supplemental charges under s 91 TCGA, would be taxed on the appellants at a rate of about 64%.

[16]Another discretionary trust was created in 2002 of which the two appellants were both beneficiaries, as were other beneficiaries of the 1969 Settlement. This is referred to in this decision notice as the 2002 Settlement. The trustees of it were Mr Bowring, the first appellant, and a Mr Ian Whiteford. It was created on 27 March 2002.

[17]Again it was accepted that the two appellants were, and were perceived by the trustees of the 2002 Settlement, as being the principal beneficiaries of this trust, and that this trust acted to implement the wishes of Mr Bowring. It was also accepted that this trust was created in order to receive the funds from the 1969 Settlement. The terms of the new discretionary trust were similar but not identical to the 1969 Settlement: but it was not suggested that the differences were in practice material. It was also accepted that, in order to avoid a breach of the rule against perpetuities, the term of the 2002 Settlement was drafted to be co-extensive with that of the 1969 Settlement. In other words, it was considered that for the purpose of the law against perpetuities, the commencement of the term of the 2002 Settlement would be seen as being on the commencement of the 1969 Settlement.

[18]In March 2002, and in execution of the flip-flop mark II planning, the trustee of the 1969 Settlement sold all the remaining chargeable assets within the trust. Some were sold to the two appellants. In particular, Mr Bowring borrowed £1,254,500 from another family trust (the Le Guet Settlement) and used this to buy from the 1969 Settlement shares in the company RFIB. At the same time Miss Bowring (acting by Mr Bowring) borrowed £625,500 from the Le Guet Settlement and paid £600,000 to the 1969 Settlement to purchase a property known as Valletta House. We refer to these two loans as the Le Guet Loans.

[19]The trustee used the trust fund which was now in cash, including the cash realised from the sale of its assets to the beneficiaries, to purchase approximately £4million worth of gilts. On 2 April 2002 the trustee of the 1969 Settlement borrowed from Butterfield Bank £3.8 million on the security of these gilts and paid £3.8 million to the trustees of the 2002 Settlement to be held on the terms of the 2002 Settlement. It was not in dispute that the reason why the assets were converted into gilts, which are fairly liquid assets, was to inject some commercial realism into the planning: otherwise the 1969 Settlement would be borrowing cash against the security of cash.

[20]The terms of the 1969 Settlement were not in dispute. The parties were agreed that the trustee had power to appoint the whole or part of the trust fund to another settlement for the benefit of one or more of the beneficiaries of the 1969 Settlement and that this power was exercised when it paid the £3.8 million to the trustees of the 2002 Settlement.

[21]The trustees of the 2002 Settlement made distributions totalling £2.4million out of the monies transferred to them by the trustee of the 1969 Settlement as follows: · 28 April 2002 £1,260,935 to Mr Bowring and £628,706 to Miss Bowring; · 15 June 2002 £58,750 to Mr Bowring and £13,750 to Miss Bowring; · 9 December 2002 £400,000 to Mr Bowring (on the understanding he would pay it to two of his cousins, £200,000 to a Miss Seward and £200,000 to a Miss Pope) · 23 February 2003 £21,000 to Mr Bowring. At the end of the tax year 2002/03, therefore, the 2002 Settlement had £1.4million remaining in assets.

[22]Mr Bowring used the distribution to him on 28 April 2002 of £1,260,935 to repay the loan to him from Le Guet Settlement together with interest; Miss Bowring used the distribution on the same day to her to repay the loan to her from Le Guet Settlement together with interest.

[23]On 2 May 2002 the trustee of the 1969 Settlement sold the gilts and repaid the loan to it from Butterfield Bank.

[24]The 1969 Settlement did not technically cease to exist at this point. A balance sheet was prepared in September 2007 which showed that in April 2007 it still had about £300 in the bank and therefore, technically, if not practically, it still existed. Who knew what?

[25]It was not in dispute that the flip-flop scheme was planned up to and including the transfer of the entire trust fund to the 2002 Settlement. It is obvious that this is so. Mr Whiteford’s memorandum (see paragraph 38 below) sets out the steps that had to be taken and the Butterfield Trust took those steps. It liquidated the assets of the trust. It used the funds to purchase gilts. It used the gilts to secure a loan. It advanced the funds loaned to it to the 2002 Settlement. It sold the gilts and repaid the loan. All these steps were planned in advance before the first step was taken.

[26]There was some mention made at the hearing that some of the minutes of actions taken by the trustee (Butterfield Trust) were not entirely accurate and it seems that this was because they were either written in advance or after the event: this only corroborates a matter, which was not in dispute, which is that the trustee planned, in advance of any of them, to take all the steps which it did take up to the transfer of the trust fund and including the repayment of the loan from Butterfield Bank.

[27]Mr Bowring and Miss Bowring (acting by her brother) were also actively involved in this planning as they assisted the 1969 Settlement to realise its assets by buying from it two assets (the shares and the house). This was also planned, including obtaining a loan from a separate trust in order to fund the purchase.

[28]The legal adviser, Mr Whiteford, who advised Mr Bowring and Miss Bowring and the 1969 Settlement, was also actively involved in implementing the planning scheme. The scheme was first proposed by Mr Whiteford, and he agreed in advance of implementation to become one of the two trustees of the 2002 Settlement.

[29]The dispute on the evidence was who knew what when and in particular to what extent the distributions by the 2002 Settlement in 2002 and early 2003 had been planned in advance and by whom.

[30]Part of this dispute was over the extent to which that the repayment of the loans to Mr Bowring and Miss Bowring by the Le Guet trust out of the funds of the 2002 Settlement was planned and when. HMRC’s case was that it was always intended that these loans would be repaid from distributions from the 2002 Settlement, and to that extent at least, distributions from the 2002 Settlement were planned before the flip-flop scheme was implemented. The Witnesses

[31]Mr Bowring did not give evidence and we were given no explanation for this. What he knew and intended we can only infer from the circumstances and the evidence of other persons.

[32]We had oral evidence from Mr Whiteford and from Mr Hodgson of Butterfield Trust. We found both to be honest witnesses and their evidence in general reliable but we accept as justified Mr Vallat’s comment that the contemporaneous documents (bar the minutes already mentioned), to the extent there were differences, were a more reliable record of events because the events took place ten years ago and their recollections after the event were to some extent coloured by the belief that it was important that the 1969 Settlement trustees did not plan the distributions made by the 2002 Settlement. Our findings

[33]Before setting out the details of our findings we note that, in view of our decision on the law as set out below, some of the disputed issues of fact, such as whether the repayment of the Le Guet loans out of the trust funds was part of the plan, were irrelevant to our decision. Nevertheless we have set out our full findings of fact on the evidence we were presented with as the case may go on appeal and these findings become relevant.

[34]Mr Whiteford was a solicitor with Clifford Chance. He was since 1978 until he retired in 2008 legal adviser to Mr and Miss Bowring. He gave advice on private client matters such as wills and trusts.

[35]Mr Whiteford was aware that by around the end of 2001 all the income in the 1969 Settlement had been used up in distributions to the beneficiaries and any future distributions to the beneficiaries would attract capital gains tax. On 8 January 2002, Mr Whiteford, having given some thought to this fact, which would result in any future distributions from the 1969 Settlement attracting a 64% CGT liability, suggested to Mr Bowring he should implement a flip-flop scheme. Mr Bowring was interested and it was agreed that Mr Whiteford would prepare a memorandum about it. It was agreed that in the meantime Mr Bowring would not ask the 1969 Settlement for any further distributions, even though he indicated to Mr Whiteford that he expected he would need a further £75,000 in the next three months.

[36]Mr Whiteford wrote a letter on 16 January 2002 to Mr Bowring advising that a flip-flop scheme would allow the trust gains to be locked into the 1969 Settlement, and in particular would allow distributions to be made from a new trust at 0% rate of tax rather than a 64% rate of tax if the distributions were made from the 1969 Settlement.

[37]We find Mr Bowring told Mr Whiteford on 15 February that he thought it likely that the whole trust fund would be distributed, and that Mr Whiteford reported to a colleague three days later that he expected further substantial distributions would be made from the fund. We do not find that that he knew of any plan to distribute a specific figure on a specific date.

[38]Miss Richardson of Butterfield Trust wrote to Mr Whiteford on 20 February 2002 asking for an explanation of the flip-flop scheme. On 28 February 2002 Mr Whiteford wrote to Miss Richardson with a copy of his memorandum entitled “CGT Avoidance Proposal”. This document said:

“It is understood that it is likely that the trustees will make further distributions to C Bowring and J Bowring, possibly up to the whole of the value of the trust fund….In view of the fact that the trustees have a duty to exercise their powers for the benefit of the beneficiaries and that one way of achieving this would be for the trustees to make payments to them in the most tax efficient way possible, it follows that the trustees would be acting properly in considering ways in which such a tax liability might be minimised or eliminated…..In view of the likelihood of further distributions to the two principal beneficiaries in the near future the scheme should be implemented before such distributions are made.”

[39]It went on to explain the scheme in more detail and pointing out that the trustees of the new trust “could” make distributions free of tax. It mentions an IHT charge that would arise and said:

“In view of the fact that it is intended that distributions should be made to the beneficiaries anyway, whether or not the scheme is implemented, the inheritance tax charge is a cost which can be regarded as a fixed cost and not one arising as a result of this scheme.”

[40]A similar comment was made when discussing the possible failure of the scheme:

“…as the intention was to make further distributions to Clive and Juliet, possibly up to the whole of the trust fund, [the IHT] charge would have been incurred in any event.”

[41]We also note that elsewhere in the memorandum it was noted that the trustees of the new trust could either distribute the funds or invest them.

[42]The flip-flop scheme required the trustees of the 1969 Settlement to liquidate the assets of the trust in order to purchase gilts. In respect of this part of the plan, and in particular the proposed purchase of Valletta House by Miss J Bowring with borrowings from another trust, the memorandum said:

“this could be done in the knowledge that a distribution to [J Bowring’s trustees] from the 1969 trust as part of the implementation of the flip-flop would be made and the borrowing would then be repaid.”

[43]On 6 March 2002, Miss Richardson asked Mr Whiteford for some amendments to be made to this document and for an invoice for Clifford Chance’s fees. The purpose of the changes was to make it clear that Clifford Chance was advising the Butterfield Trust and that the Butterfield Trust was relying on that advice: without these changes, the Trust would have instructed independent solicitors.

[44]Mr Whiteford was happy to make the changes Miss Richardson requested as he regarded himself as advising the Trustees as well as Mr and Miss Bowring, and indeed regarded their interests in this matter as identical. And, therefore, the trustees did not instruct independent solicitors.

[45]The Butterfield Trust relied on Mr Whiteford’s advice and approved the adoption of the flip-flop scheme on 7 or 8 March 2002. The matter was seen as urgent as everyone wanted it implemented before the tax year ended. And, as can be seen from the above recital of events, it was implemented just before the end of the tax year.

[46]A file note of a conversation on 19 March 2002 between Mr Whiteford and Mr Bowring about whether the new trust should be interest in possession or discretionary, records: ” [Mr Whiteford] said in view of the fact that the funds would be distributed to Clive outright fairly shortly after the trust is set up, so it may not particularly matter”

[47]Shortly after implementation of scheme, on 18 April, Mr Whiteford had a further conversation with Mr Bowring. “…we discussed the steps now to be taken in relation to the distributions from the Clive Bowring 2002 Trust. I said that was no reason why the loans to Barclay Trust should not now be paid….Given that the rate of interest was 3% over base there would seem to be little point in delaying repaying the loans….. We then discussed future distributions….” There followed a long discussion about what the original settlor’s intentions were, past distributions from the 1969 Settlement, the needs of the beneficiaries, and Mr Bowring and Miss Bowring’s desire to give a sum of money to two cousins tax free.

[48]There is then a note of a meeting on 17 October 2002. There was discussion of how to give £400K to the two cousins without a tax liability falling on the cousins. It then records:

“We then discussed whether any further distribution should be made from the trust.”

Mr Bowring said he was keen to keep funds in the 2002 Settlement to meet any CGT if the flip-flop scheme did not work. Mr Whiteford records:

“I said that I had understood that the originally (sic) intention was to pay out the trust funds fairly quickly so that the need for investment would not arise. However, in view of Clive’s present feeling that he would prefer to keep these funds in reserve, we would need to consider investment.”

[49]It was put to Mr Whiteford that actual distributions were planned before the transfer of the trust funds to the new trust but he denied this. It was put to him that it must have been planned that the loans to Mr and Miss Bowring would be repaid to the Le Guet trust out of distributions from the 2002 Settlement. Mr Whiteford said that even this was not planned in advance as Mr Bowring could have repaid the money by selling the shares and we note that there was a reference in the file notes of a possible sale of them to a third party.

[50]Our conclusion from all this is that Mr Whiteford expected that the whole or substantially the whole of the trust fund would be distributed shortly after the 2002 Settlement was established. We think, on the evidence of the conversation on 18 April, he had expected that the loans would be repaid out of the trust funds. However, we accept that that was an expectation only: there was no definite plan of which he was aware for specific distributions to be made to beneficiaries.

[51]We consider that this was his expectation because(1) Mr Bowring had said a number of things which indicated that Mr Bowring intended to ask for trust funds to be distributed;

(2) Mr Whiteford had proposed the flip-flop because he expected the trust fund to be distributed sooner rather than later; and

(3) he had not applied his mind before October 2002 to the question of investing the funds of which he had become trustee in April because he expected that the funds would soon be distributed.

[52]But once the flip-flop was proposed there was no need (and perhaps no time) 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. The Butterfield Trust

[53]Butterfield Trust (Guernsey) Limited was a trust company incorporated in Guernsey and was the sole trustee of the 1969 Settlement from 1995 until 2 April 2002, which was the date on which it appointed out the funds of the trust to the 2002 Settlement. It was at no time resident in the UK.

[54]We had no evidence from the directors of the Butterfield Trust at the time of what they knew or intended. Day to day responsibility for the affairs of the 1969 Settlement was left to an employee of the Butterfield Trust, a Miss Sue Richardson. She is no longer employed by the Trust and we had no evidence from her.

[55]We did have evidence from Mr Hodgson, who is now a director of the Butterfield Trust, although he was not at the time. He had no involvement in the affairs of the 1969 Settlement until he was brought in on 7 March 2002 to help assist with the organisation of the flip-flop scheme and in particular to arrange the necessary borrowing by the trust.

[56]We consider that the Butterfield Trust must be taken to know anything known by Miss Richardson (to whom they entrusted the day to day management of the trust) and Mr Hodgson (to whom they entrusted part of the implementation of the flip-flop scheme).

[57]We heard hearsay evidence from Mr Whiteford (recorded in a contemporaneous memorandum) that in January 2005 (nearly 3 years later) Miss Richardson had told him that she had no idea what distributions were made by the 2002 Settlement trustees after the transfer of funds and did not even known that distributions had been made shortly after the transfer of funds. She indicated that she had had no advance discussions on the point. We accept that Mr Whiteford accurately recorded what he was told.

[58]Mr Hodgson’s evidence was that he did not know of any planned distributions to be made by the 2002 Trust.

[59]We find that the Butterfield Trust was relying on the memorandum mentioned in paragraph 38 above and therefore must have expected, and did expect, as that memorandum indicated, that there would be substantial distributions out of the new trust. The whole purpose of the flip-flop scheme was to facilitate such distributions. We find that there is no evidence whatsoever to suggest that the transfer was made on condition that certain future distributions were made and as find that as a matter of law the Butterfield trust had no say in what the trustees of the 2002 Settlement chose to do.

[60]Mr Hodgson did not know how Mr and Miss Bowring funded their purchase of the assets from the 1969 Settlement and did not know that they would use distributions from the 2002 Settlement in order to repay loans they took out to enable the purchases. It is likely, however, that Miss Richardson knew that the purchase of assets was funded by loans: this is because she consented to the release of the memorandum to the Le Guet trust. We therefore find that she did know this. And her knowledge must be imputed to the trustee, Butterfield Trust.

[61]We find that the reason the Butterfield Trust chose to transfer the entire trust fund to the 2002 Settlement was because it believed this to be in the best interests of the beneficiaries as they were advised it should enable distributions to be made to the beneficiaries free of tax, and they expected substantial distributions, quite possibly to the entire value of the trust fund, to be made in the relatively near future.

[62]Mr Hodgson was involved in order to arrange at very short notice a very substantial, and somewhat unusual loan, to the 1969 Settlement. Two banks were approached. The draft proposal stated:

“…The loan is required to facilitate distributions to beneficiaries of the trust and in order to maximise the tax benefits of the scheme we would seek to obtain the highest loan to value ratio. The loaned funds would be distributed directly to the two beneficiaries and not to the trust’s bank accounts.”

[63]Mr Whiteford picked up on this in a conversation with Miss Richardson on 21 March and asked for the proposal to be corrected. The corrected version said:

“…The loan is required to facilitate distributions to beneficiaries of the trust and in order to maximise the tax benefits of the scheme we would seek to obtain the highest loan to value ratio. The loaned funds would be distributed directly to a second trust and not to the trust’s bank accounts.”

[64]The loan offers from the two banks approached (not surprisingly in view of the terms of the proposal) both indicate that the lenders were of the opinion that the loan to the trust was to assist the trust in making distributions to the beneficiaries.

[65]Mr Hodgson’s point is that the proposal did not go into detail that the lenders did not need to know: it was critical that the lenders understood that the loaned funds would be given away by the borrower as that affected credit risk. What then happened to the funds thereafter would be irrelevant to the lenders; and therefore Butterfield Trust’s trust proposal referred to distributions to beneficiaries only in the loosest sense.

[66]We find, however, that, consistently with all the other evidence, while there is no indication that the trustee of the 1969 Settlement planned particular distributions on particular days, it did expect substantially the whole of the funds to be shortly distributed by the 2002 Settlement and its reason for obtaining the loan and undertaking all the other steps in the flip-flop scheme was to facilitate tax free distributions to the beneficiaries. Conclusions

[67]What did Mr Bowring and Miss Richardson know?

[68]We find that from the outset Mr Whiteford and Mr Bowring had a clear and defined plan to implement a flip-flop mark II scheme and in particular to have the entire trust funds from the 1969 Settlement transferred into a new settlement in order for the funds to be distributed to the beneficiaries with the understanding that as a matter of law such distributions would be free of CGT to which they would be subject if the funds were distributed directly from the 1969 Settlement.

[69]We are also satisfied that from the outset it was the intention and expectation of Mr Whiteford and Mr Bowring that the funds would be distributed in the relatively near future principally to Mr Bowring and his sister but no final decision had been taken on exactly what distributions would be made and on what dates. In the event we find that rather less was distributed than originally intended when the plan was devised.

[70]It seems more likely than not that Mr Bowring (and therefore Miss Bowring) intended the Le Guet Loans to be repaid from 2002 Settlement and so we find. This is because it was not suggested that the parties considered any other source of funding to repay the loan made to Miss Bowring, and in respect of the loan made to Mr Bowring, where an alternative source of funding had been suggested, we find that the proposed purchase of the shares was vague and indefinite. Lastly, we consider that any doubt on this question of fact should be resolved against the appellants as Mr Bowring chose not to give evidence and we were given no explanation for this. As we have said we find that Mr Whiteford expected the loans to be repaid by distributions from the 2002 Settlement, although there was no plan for specific distributions on specific dates.

[71]The Butterfield Trust did not instigate or mastermind the plan but was fully aware of it and chose to play its part in it. In particular, it decided to realise the gains, to borrow the funds to trigger s 90(5) TCGA, to repay the loans, and finally (so far as the Trust was concerned) to transfer the whole trust fund to the 2002 Settlement. Mr Whiteford and Mr Bowring’s plan became its plan. It adopted Mr Whiteford’s tax avoidance planning in the expectation funds would be distributed in the relatively near future to persons who were beneficiaries of the 1969 Settlement but it did not know how much would be distributed to each beneficiary and when.

[72]We find it more likely than not that Miss Richardson (and therefore the Butterfield Trust) did not know that the Le Guet loans would be repaid out of the distributions. While we do not know what Mr Bowring told Miss Richardson, there was no particular reason why he would have informed Miss Richardson of his intention to repay the Le Guet Loans from the trust funds, as she was not his adviser, was unconnected with the Le Guet Trust, and would not be a trustee of the 2002 Settlement. Therefore, while noting that Mr Whiteford’s evidence of what she said was hearsay, we accept that both the note was accurate and that it was more likely than not to be a correct statement by Miss Richardson.

[73]The distributions which did take place were made on the decision of the trustees of the 2002 Settlement alone. We are satisfied that there was no agreement or understanding between Butterfield Trust and the trustees of the 2002 Settlement as to what distributions would be made – and indeed such an agreement would not have been legally binding. The final decisions on the dates and amounts of distributions were not taken by trustees of 2002 Settlement until after it had received the trust funds. The Law

[74]We have already recited ss 87 and 90 TCGA. HMRC’s case is that s 97(5)(a) TCGA applied to treat the appellants as receiving the distributions from the 1969 Settlement (and thus be taxed in respect of the stockpiled gains of that settlement). This appeal turns on this subsection. It provides:

“(5) For the purposes of sections 86A to 96 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.”

[75]We note in passing that the law has been since amended (by FA 2003) to block flip-flop mark II schemes but that was after the events at issue in this appeal and we need not consider it.

[76]Before turning to the crux of this case, which is s 97(5)(a) TCGA, we mention the other provisions of TCGA relevant to this appeal.

[77]S 91 causes an effective increase in the rate of tax payable under s 87. There was no dispute with regards to its application so we do not set out the full provisions here.

[78]S 97(1) defines a capital payment as a payment which is not chargeable to income tax on the recipient. Nothing turns on this provision in this appeal as it is accepted that the payments received by the appellants were capital payments. The Herman case

[79]We are not the first tribunal required to consider flip-flop mark II planning and the meaning and effect of s 97(5) TCGA. The issue came in front of the Special Commissioners (Sir Stephen Oliver QC) in the case of Herman [2007] UKSPC 609. In that case (paragraph 12), as in this case, it was common ground that the implementation of flip-flop mark II planning was effective to prevent the realised gains in the original settlement being transferred into the new settlement. The question was (and is) whether s 97(5)(a) applies to treat the beneficiaries who received distributions from the new settlement as receiving capital payments from the original settlement.

[80]In that case the planning scheme was first proposed to the beneficiaries by the trustee of the original settlement (a Guernsey company), Orbis Management Ltd. All the advice in respect of the planning was given, and to a large extent all the implementation was carried out, by Orbis Management Ltd and what appears to have been a company connected to it, Orbis Taxation Services Ltd. The decision appears to consider them as acting as a single entity and refers to them as “Orbis”.

[81]The conclusion Sir Stephen reached on the facts and law was that Orbis planned and implemented, with the beneficiaries’ agreement and full knowledge, a series of transactions, all of which were related, the outcome of which was intended if not preordained:

“(43)….That outcome was the release of the funds originating from the [original] settlement to Mr Herman and Mrs Herman absolutely. To conclude otherwise would, I think, be shutting one’s eyes to the obvious. (44) For all those reasons I have concluded that the amounts transferred to them from the [new] settlement …were received by Mr and Mrs Herman indirectly from the trustees of the [original] settlement for the purposes of s 97(5)(a). (44) I dismiss the appeal…..”

HMRC’s submissions

[82]Mr Vallat says that as a matter of ordinary language all the payments were received directly from the 2002 Settlement and indirectly from the 1969 Settlement. It is HMRC’s case that this does not give rise to a double charge to tax but rather the real source of the payment should be identified.

[83]It is HMRC’s case that the real source of the payment was the 1969 Settlement and not the 2002 Settlement because:(a) The trustees of the 1969 Settlement knew that distributions were intended; and/or

(b) Even if they did not, Mr Whiteford did, and the trustees of the 1969 Settlement acted on his advice; and/or

(d) In all the circumstances the distributions were really from the 1969 Settlement.

[84]His primary case is that the Butterfield Trust (via Miss Richardson and Mr Hodgson) knew of the broad plan to make the distributions from the 2002 Settlement; his secondary case is that if they did not, then Mr Whiteford did.

[85]Mr Vallat considers that Mr Prosser’s case amounts to saying that the question is who really controls the money and it is only where the immediate payer is an intermediary for someone else that he can be disregarded. But, says Mr Vallat, if Mr Prosser is right this would make “indirectly” otiose. Appellant’s submissions

[86]The question is whether Mr and Miss Bowring received the distributions made to them by the trustees of the 2002 Settlement from the trustee of the 1969 Settlement as s 97(5) only creates liability where the taxpayers “receive it from” the 1969 Settlement “directly or indirectly”.

[87]The appellants’ case is that, where money is received from one party (in this case the 2002 Settlement) it can only be said to be received from another party (the 1969 Settlement) if the 2002 Settlement was acting as an intermediary for the 1969 Settlement.

[88]And where, say the appellants, the 1969 Settlement has given the money to the 2002 Settlement with no strings attached, and in particular no obligation to make distributions to anyone of any amount, and the 2002 Settlement decides to make distributions of its own choice, those distributions can not be said to be directly or indirectly “from” the 1969 Settlement even if the funds originated with that Settlement. The 2002 Settlement has acted independently of the 1969 Settlement so, say the appellants, s 97(5)(a) cannot apply. Indirectly?

[89]What does “from … directly or indirectly” mean?

[90]We think the phrase should be construed purposively. We find it was clearly intended to be wide in meaning. Its context is legislation intended, in broad terms, to tax beneficiaries on gains incurred in respect of property held by non resident trusts but only to the extent that the beneficiaries benefit from the gain. In broad terms, if it were not for this legislation, it would allow a person to benefit from the proceeds of property on which gains have been realised yet not have any CGT liability because the gains were realised while the property was held in offshore trusts. The context of s 97(5) is therefore anti-tax avoidance.

[91]And s 97(5) is itself clearly intended as an anti-avoidance provision within this general anti-tax avoidance context: it is intended to catch the many different ways in which a person could obtain a benefit from a trust other than by a simple transfer of money from the offshore trustees. To recap, it provides:

“(5) For the purposes of sections 86A to 96 and Schedule 4C a capital payment shall be regarded as received by a beneficiary from the trustee 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.”

[92]We are concerned with paragraph (a). But it must be seen in its context so we look at s 97(5)(b) and (c) as well.

[93]Paragraph (b) suggests that the provision is intended to a catch a situation where the trust money is paid to a third party who would then have a contractual obligation to provide some kind of service to the beneficiary. Where this has happened, the beneficiary will have lost the equitable interest in the property but gained the benefit of the performance of a contract. An example might be where the trust funds are used to pay the rent on a house for the beneficiary to live in or to pay his university fees or to buy him a share in a partnership.

[94]Paragraph (c) catches a situation where money is paid to a third party at the beneficiary’s direction. There is no contract with this third party. The trustee has chosen to give the trust funds to someone at the option of the beneficiary. It does not seem to be a requirement of (c) that the third party should actually use the money for the beneficiary’s benefit, although it is difficult to see that a trustee could make such a payment unless it was understood the third party would use it for the benefit of the beneficiary or the third party was also a beneficiary of the same discretionary trust. An example of a situation in (c) might be a payment to the spouse of the beneficiary.

[95]Section 97(5) does not expressly contemplate the case of money held in one trust being passed to another trust. But it was clearly intended to be interpreted widely. In (a) and (b) the expression “directly or indirectly” is used, and clearly (b) and (c) are intended to catch situations where the beneficiary only indirectly benefits from a payment made by the trustee. Moreover, while this case was argued on s 97(5)(a), one way of reading s 97(5) is that (b) and (c) apply too:(1) So far as (b) is concerned, in paying the funds of the 1969 fund to the 2002 Settlement, the trustee of the 1969 Settlement was applying it for the benefit of Mr and Miss Bowring, as not only were they included in the class of beneficiaries of the 2002 Settlement, it was virtually certain that the trustees of the new trust would continue the 1969 Settlement treatment of Mr and Miss Bowring as the main beneficiaries as one of the trustees was Mr Bowring;

(2) So far as (c) is concerned, it could be said that the trustees of the 2002 Settlement received the trust fund at the direction of Mr Bowring, in that Mr Bowring made it quite clear that he wished the Butterfield Trust to participate in the avoidance scheme and pass the trust funds to the 2002 Settlement.

[96]These two points were not argued, but the point we are making is that s 97(5) was intended as a broad anti-avoidance provision and “from ….directly or indirectly” should be interpreted in that context.

[97]Moreover, while (a) and (b), if the word “indirectly” is ignored, appear to only apply in a case where the trustee has made the decision as exactly how and when the beneficiary is to benefit from the funds, (c) clearly contemplates a situation where the choice of how and the extent to which a beneficiary will benefit from the payment has passed from the trustee to a third party. Therefore, when the word “indirectly” is added back in, even in the case of (a) and (b) it is not obvious that they were intended to be limited to situations where the trustee would decide exactly how and when the beneficiary would benefit from the trust fund. The funds must be “from” the trustee but not necessarily immediately from the trustee. The trustee must be the root of the funds or benefit.

[98]The context alone is enough to suggest to us that the appellant’s view of “indirectly” is too narrow. The appellant suggests “indirectly” could only refer to an intermediary for a principal (whether or not in the legal sense), where the original trustee has determined, if it does not control, the ultimate payment to the beneficiary. We think the context means it has a broader meaning than this. S 97(5) clearly anticipated a distribution via a third party.

[99]But is it enough for s 97(5)(a) to apply merely for the funds ultimately distributed to Mr and Miss Bowring to have originated in the 2002 Settlement or is more required? Is it merely a tracing exercise?

[100]Both parties are agreed that the question whether “he receives it from [the trustee of a settlement] directly or indirectly” is not solely a question of tracing. Sir Stephen in Herman also did not consider it to be solely a question of tracing.

[101]Sir Stephen considered there were three signposts to determine whether s 97(5)(a) applied:

“(21)….An obvious signpost will be the existence of a plan, if there is one….The second signpost is to analyse the trust law and determine whether the [new trust] ‘served as a vehicle to receive and continue the act of bounty effected by’ the trustees of [the original settlement]…. The precise means by which the scheme was implemented will, in addition, be relevant to the question whether there is sufficient linkage to make the payments ‘indirect’ receipts from the trustees of the [original settlement].”

[102]Mr Prosser criticised Sir Stephen’s decision: he said he was wrong to have substituted his three signposts for the statutory test. He particularly criticised the second signpost referred to by Sir Stephen which, because of Sir Stephen’s use of the word “tracing” in an earlier paragraph, he considered related to tracing back the funds. He says Sir Stephen has misunderstood or at least misapplied what Lord Walker said in West v Trennery

[2005] UKHL 214

and from which he quoted in the above passage. To consider this we have to look at that case.

[103]We comment that it is clear that Sir Stephen’s second signpost was not merely about tracing funds: it was looking at whether in practice the second trust was merely a continuation of the first trust. If A gives funds to B who gives them to C, the funds can be traced from C back to A. But that does not mean C’s receipt was necessarily “from” A, either directly or indirectly. It is more than a tracing exercise. We consider the second signpost in more detail. Where the second trust continues the act of bounty of the first The Trennery case

[104]This case also involved a type of flip-flop scheme to mitigate tax liability; although it involved different statutory provisions, at root the scheme had the same fundamental basis as the other flip-flop schemes we have already outlined. Trust 1 would advance borrowed funds to Trust 2 rather than the trust property itself.

[105]In summary, Mr Trennery wished to avoid paying higher rate CGT on some shares he wished to sell. So he gave the shares to a trust (of which he later ceased to be a beneficiary). This trust, the original trust, borrowed funds on the security of the shares. It gifted the borrowed funds to a new trust, of which Mr Trennery was also a beneficiary. It then sold the shares and repaid the loans.

[106]The question was whether Mr Trennery had to pay CGT on the sale of the shares at the 40% rate applying to individuals who were interested in a trust or at the rate applying to trustees (25%).. The question was whether Mr Trennery (a higher rate tax payer) had an interest in:

“any property … comprised in the settlement or any derived property is … payable to or applicable for the benefit of the settlor; or (b) the settlor…enjoys a benefit deriving directly or indirectly from any property which comprised in the settlement or any derived property.”

[107]The Lords considered that Parliament intended to catch income or capital which directly or indirectly represented proceeds from assets held by the first settlement. As Mr Trennery was a beneficiary of a trust which held funds indirectly generated from the first settlement (the money loaned to the first settlement on the security of the trust property), higher rate tax had to be paid.

[108]The criticism of the Herman case is that it focuses on one small point in Trennery and, in Mr Prosser’s view, misunderstands it. At paragraph 41 Lord Walker is considering whether the funds in Trust 2 were derived from the shares held in Trust 1. He makes the comment that the two trusts were separate but then qualifies this by a paragraph in parenthesis.

[109]That paragraph was as follows:

“[41] ….(to state, as the respondents’ printed case does, that it was an entirely separate settlement might be said to overlook the effect of the rule against perpetuities, as explained by this House in Pilkington v IRC [1964] AC 612…; the trust law analysis is that the second settlement served as a vehicle to receive and continue the act of bounty effected by the first settlement, with the rule against perpetuities acting as a sort of umbilical cord between the two settlements; the fact remains, however, that it was a separate settlement for CGT purposes.)”

[110]Mr Prosser’s point is that Lord Walker, although he made this comment, was clearly not relying on the connection between the trusts created, or at least recognised, by the rule against perpetuities, in reaching his conclusion. It was an aside and Lord Walker treated the two trusts as entirely separate for CGT purposes. Yet Sir Stephen regarded the connection between the trusts, recognised by the rule against perpetuities, as one of his signposts indicating that the distribution from the second trust was indirectly from the first trust. Mr Prosser is really saying that one of Sir Stephen’s signposts is misconceived.

[111]It is clear that there was a straightforward slip in Sir Stephen’s adoption of Lord Walker’s words. Lord Walker referred to the act of bounty “effected by the first settlement”. Sir Stephen referred to the act of bounty effected by the trustees of the first settlement. The act of bounty was, of course, the act of bounty of the settlor in effecting the first settlement. And Sir Stephen would be well aware of this.

[112]Correcting this obvious slip and reading it as a reference to the settlor of the first settlement, is this signpost of Sir Stephen’s to be criticised on any other grounds? Was he wrong to rely on tracing and the law against perpetuities? The law against perpetuities

[113]Does the law against perpetuities have any relevance here even if only by analogy? Lewin on Trusts , 18 th edition at Section 35-90 summarises the rule on perpetuities where settled property is re-settled, as follows:

“A power of advancement (including the statutory power) is to be treated as a special power of appointment for the purposes of the rule against perpetuities. If it is exercised so as to take property out of the settlement and settle it on new trusts, the perpetuity period for those trusts runs from the date of the original settlement, not from that of the advancement.”

The author cites Pilkington (see the citation from Trennery) as the authority for this proposition.

[114]Lord Radcliffe gave the leading judgment on perpetuities in Pilkington, saying, in a case which involved the legality of trustees of one trust creating a new trust on different terms for one of the beneficiaries:

“…..When one asks what person can be regarded as the settler of Miss Penelope's proposed settlement, I do not see how it is possible to say that she is herself or that the trustees are. She is the passive recipient of the benefit extracted for her from the original trusts; the trustees are merely exercising a fiduciary power in arranging for the desired limitations. It is not their property that constitutes the funds of Miss Penelope's settlement; it is the property subjected to trusts by the will of the testator and passed over into the new settlement through the instrumentality of a power which by statute is made appendant to those trusts. I do not think, therefore, that it is important to this issue that money raised under a power of advancement passes entirely out of the reach of the existing trusts and makes, as it were, a new start under fresh limitations….. I think that the important point for the purpose of the rule against perpetuities is that the new settlement is only effected by the operation of a fiduciary power which itself “belongs” to the old settlement.”

The decision was that the law of perpetuities meant that the new trust had to have the same limit in time as the old trust: it had to be treated as starting at the time the old trust started. Re-settling the trust funds did not re-start the clock.

[115]This application of the rule against perpetuities could be seen as a common law anti-avoidance rule based on public policy. Its effect is to prevent avoidance of the rule against perpetuities, by making void the settlement of trust property on new trusts, the time period of which, if measured from the start date of the original settlement, would exceed the time period for which property is permitted to be held on private trusts. Trust property can therefore be re-settled, but the new trust will only be valid if its time period, as measured from the start of the original settlement, does not breach the rule against perpetuities. This law treats the second trust, for these purposes, as being the same trust as the original trust.

[116]It treats the second trust as if it were the first trust on the basis that the funds trace back to the first trust: it is irrelevant that the terms of the second trust are different. The key thing, according to Lord Radcliffe, was that the real settlor of the funds was the settlor of the original trust and that the second trust could only be created by operation of a trustee power which was (or was treated as) granted by the original trust.

[117]In Trennery, Lord Walker in parenthesis was recognising a similarity with the position on the facts of that case where property had been transferred from one trust to another, not in an attempt to avoid the rule that a trust cannot exist in perpetuity, but instead to avoid tax liability.

[118]There is nothing to suggest Sir Stephen misunderstood the point Lord Walker was making. On the contrary. The difference between them is simply that Lord Walker made the point in parenthesis and it seems he was noting the similarity but not relying upon it to reach his conclusion. In Herman, Sir Stephen relied on the point. Conclusion

[119]Mr Prosser criticises Sir Stephen’s decision because it refers to tracing the funds back to the original settlement. We find these criticisms ill-founded. While Sir Stephen used the word “traced” (eg paragraph 18) it is quite clear that he is not carrying out a simply tracing exercise. In paragraph 21 he asks the question whether the distributions “can properly be linked” to the transfer from the original trust to the new trust so that the original trust was their indirect source.

[120]We think it obvious that a beneficiary of a second trust could not be said to receive the distribution “from [the trustee of another settlement] directly or indirectly” unless the distribution comprises funds that derive from the settlor of the original settlement. But it is also the case that the mere facts that the funds originated with the first settlement is not enough. There must be an additional element. Sir Stephen’s second signpost was that the second trust effectively continued the bounty of the first trust.

[121]We think that this must be right. 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 anti-avoidance 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 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](We note that Parliament referred in s 97(5) to distributions from “the trustee” whereas the distribution could perhaps be better described as coming from the original trust, or even the settlor himself, in that the trustee performs no act of bounty as it has no beneficial interest in the funds. However, we do not think anything should be read into this: trusts have no legal personality separate from their trustees. What Parliament meant to capture in s 97(5) was any case where the beneficiary receives benefit from trust funds, irrespective of the identity of the trustee, yet clause s 97(5) had to reflect the position that legally the funds are transferred by the trustee of the trust.) 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. Application to facts of this appeal

[127]We find that the funds distributed to Mr and Miss Bowring by the trustees of the 2002 Settlement can clearly be traced back to the 1969 Settlement. The 2002 Settlement was created under a power in the 1969 Settlement trust deed. Further, although both were discretionary trusts, Mr and Miss Bowring were in practice treated as the principal beneficiaries of both the 1969 and 2002 Settlements. The term of the 2002 Settlement was co-extensive with that of the 1969 Settlement. Nothing changed with the creation of the 2002 Settlement other than the identity of the trustees and some changes in the terms of the discretionary trust which were in practice of no importance. Not only that but one of the trustees of the 2002 Settlement was Mr Bowring: who was the beneficiary in accordance with whose wishes the trustee of the 1969 Settlement was accustomed to make distributions. Therefore, the 2002 Settlement was, viewed realistically, essentially the same as the 1969 Settlement.

[128]The distributions to the Bowrings by the 2002 Settlement trustees could not and would not have taken place except for the transfer of funds by the trustee of the 1969 Settlement to the 2002 Settlement. The distributions from the 2002 Settlement were, we find, indirectly from the trustee of the 1969 Settlement. We point out that were we to come to any other conclusion, it would lead to the bizarre result that a distribution would be within s 97(5) if all that happened was that the trustees of the trust resigned their position and new trustees were appointed because the settlement would be seen as continuing, but if they were to transfer the funds to new trustees with a new deed of settlement, then a distribution from the new trustees would not be caught by s 97(5).

[129]That conclusion is sufficient to dispose of this appeal but we do not leave our decision here. We go on to consider further matters. Firstly, we consider that if we are wrong to consider Sir Stephen’s second signpost and the question of whether the second trust is essentially the first trust sufficient grounds for the application of s 97(5)(a), whether s 97(5)(a) would apply to this case and in particular whether the existence of a “plan” (Sir Stephen’s first signpost) would make a difference and if there was a sufficient plan in this case.

[130]Secondly, we consider the objections the appellant puts forward to a wide interpretation of s 97(5)(a) and in particular(a) The suggestion that it must have been intended to have a narrow interpretation as other provisions in this group of sections make it clear that Parliament only anticipated it applying when the trustee existed at the time of the distribution;

(b) A wide interpretation could cause double taxation and this cannot have been intended by Parliament. Must there be a plan?

[131]As we have said, the appellants’ case is that a distribution is only “from” a trustee if that trustee decided that that particular distribution would be made. Mr Prosser’s explanation of the use of “indirectly” was to capture situations where the trustee made the decision, but passed the funds through an intermediary.

[132]Therefore, he considers that Sir Stephen was wrong to decide Herman on the basis that the trustees of the original trust had a plan to distribute the trust fund: as at the point of the distribution the funds had passed beyond their control, the original trustees did not and could not control whether the planned distribution was made or not.

[133]But this gives “from” a very narrow meaning: it means restricting it to cases where the trustees of the first settlement determine and control exactly when the distribution is made. We have already said that s 97(5) was not intended to have such a narrow meaning. In particular, we have already noted that s 97(5)(c) on its face contemplates the funds being distributed to a third party and going beyond the control of the trustee yet nevertheless, because it was done at the direction of the beneficiary, being for his benefit. So we do not think “indirectly” imports the concept of retention of control. If control was required, the words “directly or indirectly” would be superfluous as “from” would be sufficient.

[134]Therefore, if “from” does require the trustees of the first settlement to plan and intend the distribution, it does not require them to go further and actually control the distributions.

[135]In our view it is enough, if there must be a plan, for that plan to be to facilitate distributions. We have already said we think that it is enough, if the distributions from the second trust are made out of funds originally held by the trustee of the first trust where the second trust viewed realistically is a continuation of the first trust, for them to be “from” that trustee; this must even more so be the case where the transfer of the funds from one trust to another was made by the trustee with the intention of distributions being made.

[136]If a plan is a requisite (and we have said we do not think it is), then the plan must be the plan of the trustee of the original settlement. He stands in the shoes of the settlor. He controls the funds. Application to facts of this case

[137]There was a plan to facilitate distributions and in particular to enable distributions to be of the funds comprised in the 1969 Settlement but free of tax liability that would otherwise ensue if the trust distributed directly to the beneficiaries. At the time of execution of the plan, there was no plan for specific distributions to be made on specific days, but there was a general expectation that distributions probably to the whole value of the trust fund would be made within a relatively short time scale.

[138]The plan was suggested by Mr Whiteford. It was adopted and pursued by Mr Bowring. It was also adopted and implemented by the Butterfield Trust. The role of the Butterfield Trust was reactive in the sense that they did not initiate the plan. But this is completely irrelevant.

[139]It adopted the plan. It made an active decision to participate. It took all the necessary steps to implement the flip-flop mark II scheme, such as realising assets, borrowing funds and advancing the loaned funds to a new trust. It did all this with the plan and object of facilitating the distribution of the funds held in the 1969 Settlement free of tax to beneficiaries of the 1969 Settlement.

[140]We mention in passing that if we are wrong, and a plan would only be relevant if the 1969 Settlement trustees had planned the actual distributions, down to the amounts and the dates, which were actually made to the beneficiaries, then on the facts there was no such a plan. The Butterfield Trust did not know exactly how much would be distributed to whom and when; at the date the funds were transferred no one knew this. Sir Stephen’s third signpost

[141]Mr Prosser criticises Sir Stephen’s third signpost, which is whether there was sufficient linkage, on the basis that it is vague. And perhaps it is. Our view, expressed above, is that it is sufficient that the funds of second trust can be traced back to the funds of the second trust, at least where in essentials the second trust is realistically to be viewed as effectively the same as the first trust; and if we are wrong, then it is certainly sufficient for s 97(5)(a) to apply if the funds can be so traced and the transfer of the funds from the original to the new trust was intended by the trustees of the first trust in order to facilitate distributions.

[142]We do not see the need for any other linkage.

[143]But we revert now to the two main reasons why both parties considered that Parliament could not have intended so wide an interpretation to be given to s 97(5) as the one we have given:(a) The need for the trustee to exist;

(b) Double taxation. Capital payments legislation – trustee must exist

[144]The appellant says their analysis of s 97(5)(a) is supported by the capital payments legislation. This presupposes that the trust exists in the year in which the beneficiary receives the capital payment. Therefore, even in the case of an indirect receipt under s 97(5)(a) the trust must exist when the distribution is made to the beneficiary. So, says the appellant, in an “indirectly” case, the distributor must be an intermediary of the trustee of the original trust because otherwise it would not matter if the original trust still existed or not.

[145]Mr Prosser refers to s 87(4) set out above (paragraph 3) where it says: “…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 ….” (Mr Prosser’s emphasis)

[146]Mr Prosser’s case is that it is obvious that the trustee must exist for liability to tax arise because this is what the section says. If someone is dead, or, if corporate, has been wound up and struck off the company register, that person or company could not do anything and in particular could not make a capital payment.

[147]So, on his scenario, if a flip-flop scheme was implemented, Trust 1 immediately wound up and the corporate trustee struck off the register, and Trust 2 did not make any distributions until the following tax year, s 87 could not bite. Even if the distribution was planned by Trust 1 in advance and Trust 2 had promised to make it, says the appellant, the distribution could not be “from” Trust 1 as at the time of the distribution, Trust 1 did not exist.

[148]Therefore, Mr Prosser argues, “from” must have a narrow interpretation as a wide interpretation would lead to this absurdity of having a non-existent trust doing something. So s 97(5)(a) could only apply if the second trust was an intermediary of the first.

[149]We do not agree with this for a number of reasons.

[150]Even though Mr Vallat agrees with Mr Prosser’s view that the original trust must still exist at the date of the distribution, we do not. There is no reason why Parliament would not have intended to tax in the scenario outlined by Mr Prosser. The interpretation Mr Prosser argues for would let the UK resident beneficiary off the hook on the technicality of the dissolution of the trustee and we do not think this could have been Parliament’s intention. Mr Prosser suggests that the legislation was worded the way it was because it was seen as important that the trustee exist in the year of receipt. We do not agree that that was in the legislator’s mind.

[151]Further, there is no need to relate “in that year” to the phrase “from the trustee”. Section 87(4) could as easily be read as follows: “…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 ….” (our emphasis)

[152]The requirement “in that year” properly read actually relates to the whole of the phrase “receive capital payments from the trustees”. The receipt must be in that year, and the receipt must be from the trustees, but literally read there is no requirement that the trustees exist at the time of the receipt.

[153]Mr Prosser is eliding two concepts: he is stating that because (he says) a distribution from the trustee could only occur if the trustee was in existence, then it necessarily follows that the trustee must be in existence at the time of receipt.

[154]Yet it does not follow that a person is necessarily alive when a distribution from them is received. An analogy would be a person receiving a letter from someone who is dead: this is unusual but far from impossible. It only requires the sender to be alive when the letter is sent: not when it is received. The letter is still from the dead sender. S 87(4) which requires the receipt (from the trustees) to occur in the year in which it is taxed: it does not in so many words require the giver to still exist in that tax year and we do not see why we should read into it such a requirement.

[155]So, giving s 87(4) a purposive interpretation, rather than a narrow one not required on the face of it and inconsistent with the anti-avoidance intention of s 97(5), we do not agree that the trustee needs to be in existence or alive when its distribution is received.

[156]In any event, even if we agreed that the trustee had to exist in the year of receipt, we would not agree with Mr Prosser that it necessarily follows from that that Parliament intended to limit “indirectly” to cases where a person acted as an intermediary. If that had been its intention, there would have been no need for the word “indirectly” as a receipt from a person via that person’s intermediary (in the sense of agent) would in normal language be directly from that person.

[157]In passing, we note for completeness that the appellant does not directly rely on s 87(4) because it is accepted, and we find, that the 1969 Settlement was still in existence in 2007, long after the distributions at issue in this case.

[158]In conclusion, there is nothing in s 87(4) that would suggest that s 97(5)(a) could not treat a distribution made by a second fund, out of funds transferred to it on trust from the first fund, as a distribution by the first fund, whether or not that first fund was still in existence at the date of the distribution. Double counting and double taxation

[159]Mr Prosser’s second argument in favour of a narrow interpretation of s 97(5)(a), as we understand it, is because otherwise, he says, there would be the possibility of double taxation and this cannot have been intended by Parliament.

[160]The purpose of s 90 was to transfer the gains from an original settlement to a new settlement, so that the taxing provisions of s 87 could not be avoided by transferring the funds from the original trust to a new trust. S 90 is predicated on the basis that the tax charge on the beneficiaries will arise because the new trust makes a distribution to them. Yet if s 97(5) is given a wide interpretation, this means that tax law is treating the beneficiaries as receiving a distribution from both the original and new trust at the same time. This would lead to double taxation: the beneficiary could be taxed twice on the same distribution. Parliament, says Mr Prosser, cannot have intended double taxation, particularly where, with an effective 64% rate of tax, as in this case, such double tax would actually exceed the amount of the distribution.

[161]If “indirectly …. from” is limited to cases where one trust is an intermediary for another, says Mr Prosser, there is no risk of double taxation. The payment is not “from” the trust which acts as an intermediary.

[162]It is irrelevant, says the appellant, that there is no risk of double taxation in this particular case, because a wide interpretation of s 97(5)(a) could lead to double taxation in cases where there is no tax avoidance motive or scheme.

[163]We agree that, as in Herman, there is no risk of double taxation in this case. This is because s 90 is disapplied so that the gains in the first trust are not transferred to the second trust and the second trust has no gains of its own and section 87 cannot apply to the second trust as it had UK resident trustees. The view in Herman

[164]The double taxation point was considered by Sir Stephen Oliver in Herman. The appellant’s counsel had put forward examples of double counting which are in an Appendix to Sir Stephen’s decision. He said at paragraph [21] that:

“…it is not beyond the realms of purposive construction that section 87(5) might apply to prevent the double counting problems.”

[165]Section 87(5) is set out in full above but we repeat it here:

“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.”

[166]Mr Prosser says this is not intended to deal with double counting. It is intended to ensure that, however large the gains realised by the trustees of a single settlement are, they are only attributed to the beneficiaries to the extent that the beneficiaries actually receive distributions.

[167]Mr Prosser gave us an hypothetical example. This example involved a transfer of funds from Trust A to Trust B, and a distribution from Trust B to a UK resident beneficiary. Trust B made a gain in the year of the distribution; Trust A made a gain in the subsequent year. The gains were in excess of this distribution.

[168]The beneficiary has received a distribution “from” Trust B. The effect of s 87(4) is that the beneficiary is taxed on the distribution from Trust B as it applies to gains realised in the year of distribution. S 87(5) applies to limit his tax liability to the amount of the distribution actually received from Trust B.

[169]But, on a wide interpretation of “from” in s 97(5)(a), the following tax year the beneficiary is also, says Mr Prosser, liable to be taxed on Trust A’s gains realised in that following year because s 87(4) applies where capital payments were received “in any earlier year.” While s 87(5) would apply to limit this tax liability to the amount of the distribution, s 87(5) cannot be used across tax years, says Mr Prosser, to prevent double taxation because it refers to the attribution under s 87(4) and the attribution in s 87(4) is of gains in a current year.

[170]Mr Vallat agrees with Mr Prosser on this.

[171]We find that while the gains can only be attributed in a current year and to be liable B must have received a capital payment in that or in an earlier year, there is nothing on its face to limit the application of s 87(5) in time. In particular it does not say that the attribution under (4) shall not exceed capital payments received by B in that year. It is a very general provision.

[172]S 87(5) does refer to “attribution” rather than “attributions” but it clearly means attributions in the plural because it refers to “beneficiaries” and “payments” in the plural. So the drafter was clearly thinking of more than one attribution.

[173]However, while “attribution” must have been used to refer to more than one attribution, Mr Prosser’s point is that subsections (4) and (5) should be read together and s 87(4) is referring to attributions to a number of beneficiaries but from a single settlement in a single year of assessment, and, as (5) is intended merely as a qualification to (4), and in particular to prevent attributing the same gain to more than one beneficiary, it must therefore be read as referring to attributions from a single settlement in a single year of assessment.

[174]Further, s 87(6) is obviously intended to prevent double taxation by leaving out of account capital payments to which gains have been attributed in earlier years. So s 87(5) is intended to prevent double taxation which might otherwise arise as s 87(4) might be read as taxing different beneficiaries on the same gains incurred by the same trust in the same year of assessment, and s 87(6) is intended to prevent double taxation which might otherwise arise under s 87(4) if it was applied to the same person on the same gains in different years. Subsection (6) would be unnecessary if subsection (5) had been intended to apply across different tax years.

[175]Read in its context, therefore, there is something in Mr Prosser and Mr Vallat’s interpretation of s 87(5) as being limited in scope, even though in isolation it could be read as being of more general scope.

[176]But legislation should be read purposively which means that s 87(5) should not be seen just in the context of s 87 as a whole, but in the context of ss 86 to 97 as a whole. Parliament must be assumed not to have intended double taxation and the existence of subsections (5) & (6) shows that it quite clearly did not intend s 87 to result in double taxation, whether by two persons both being taxed on the same gain or one person being taxed more than once in respect of the same distribution.

[177]While we agree that the drafter of s 87(5) did not have in mind the possibility that a single distribution might be seen as taxable twice on a single beneficiary because of the provisions of s 97(5) might lead to it being received directly from one trust and indirectly from another trust, it is clear that he was giving effect to Parliament’s intention that beneficiaries should only be taxed to the extent distributions were actually received.

[178]So we agree with Sir Stephen that s 87(5) should be given a purposive interpretation. It should be applied literally and without reference to s 87(6): the effect is that any attributions to any person under s 87(4) cannot exceed the amounts of capital payments received by them from any trust in any year.

[179]In effect we have the choice of giving some part of the statute a meaning that it was not intended to bear. Mr Prosser wants us to give a very narrow interpretation to s 97(5)(a) and limit it to payments via intermediaries, yet the context makes it clear it was intended to have a much wider meaning.

[180]Alternatively, we can read s 87(5) literally, ignoring its immediate context, and treat it as applying across tax years so that we can give s 97(5)(a) the meaning it was intended to have and still avoid double taxation.

[181]As Parliament quite clearly intended to tax, although just the once, in the scenario outlined, we much prefer the second option. We consider that s 87(5) would apply in the scenario outlined by Mr Prosser and prevent double taxation in those cases where it might be an issue (but, as we have noted, not in this one). HMRC’s solution to the risk of double counting

[182]Mr Vallat’s solution to the double counting issue is superficially similar to Mr Prosser’s in that they both agree that the legislation should be read so that the payment is only “from” one trust or the other trust. Where they differ is on what “from” means. Mr Prosser’s view is set out in more detail above but in summary is a question of who the payment is really from: it is really from the trust which distributed the money to the beneficiary unless it was acting as an intermediary for another; Mr Vallat’ view is that the answer is determined by whether there was a plan for the money to be passed to the second trust in order to distribute it. Taken to extreme Mr Vallat’s view is that there isn’t a ‘transfer’ in the sense intended by s 90 if the funds were given by the original trust to the new trust with the intention of them being distributed.

[183]Mr Prosser is critical of Mr Vallat’s view. The trustees of the 2002 chose to make the distributions at issue in this case: how can the receipt of the distributions not be “from” the 2002 Settlement? How can the old trustees having a plan make it less of a direct receipt from the new trustees? Further, the premise of s 90 is to attribute the original trust’s gains to the new trust because the law would treat a distribution out of the new trust as being “from” that trust. And as s 90 is an anti-avoidance provision, the archetypal situation in which it must have been intended to operate was when there was a plan to avoid tax by transferring the property to a new trust. To accept Mr Vallat’s view of the law would be to deprive s 90 of all effect.

[184]Mr Prosser’s criticism is, we think, well founded. It is quite clear that s 90 was intended to operate even if there was a plan to avoid tax by transferring the money to the second trust prior to distributions taking place although we note that an interpretation that the distribution is not “from” the second trust in these circumstances would not result in loss of tax as the tax would be due under s 87 rather than s 90.

[185]In conclusion, we reject both Mr Prosser’s and Mr Vallat’s solution to the double counting issue and adopt Sir Stephen’s, as the only one which permits us to fully give effect to Parliament’s intentions in ss 86 to 97. The position of the cousins

[186]We have determined the appeal against the appellants at least in so far as they received and kept distributions from the 2002 Settlement. But what of the payment of £400,000 to Mr Bowring which he passed on to his two cousins? HMRC’s assessment included tax on this £400,000: was this right? If it was wrong, they will be unable to tax the two cousins as they are out of time to do so.

[187]It was accepted that Mr Bowring wanted two cousins to receive funds from the trust and he wanted them to receive the funds free of tax. Both Miss Seward and Miss Pope were beneficiaries of both the trusts and the trustees of the 2002 Settlement could have made the distribution to them directly. The Trustees of the 2002 settlement distributed £400,000 to Mr Bowring on 9 December 2002 on the understanding that he would (as he did) then give £200,000 to each of his two cousins. The distribution was made in this indirect manner because Mr Bowring wanted to ensure his two cousins received £200,000 net: he wanted to protect them against tax liability. So he thought that by receiving the distribution himself he would be the one liable to pay any tax that arose if the flip-flop mark II planning was unsuccessful.

[188]Mr Prosser’s position was that this was actually an example of a distribution made indirectly by a trust. The distribution to the two cousins really came from the 2002 Settlement but directly it was from Mr Bowring. Mr Bowring acted as an intermediary (even if not in the legal sense) for the 2002 Settlement. Therefore the tax should have fallen on the cousins and not on Mr Bowring.

[189]His view would be that a wide interpretation of s 97(5) leads to double taxation: Mr Bowring would be liable to tax on the £400K as he received this directly from the 2002 Settlement and the two cousins between them would also liable to tax on £400,000 as they received this indirectly from the trust.

[190]In fact Mr Vallat’s gloss on this interpretation would also lead to double taxation. He would have us ask whether the distribution to the cousins was “really” from Mr Bowring or the trust. It would be “really” from the 2002 Settlement as there was a plan. But Mr Bowring would also have really received the funds from the 2002 Settlement. And Mr Bowring and the cousins could also be said to have indirectly received the funds from the 1969 Settlement.

[191]We think both counsel are wrong. We find that Mr Bowring is liable to tax as he received the distribution indirectly from the 1969 Settlement for all the reasons given above.

[192]The cousins however could not be liable to tax on the same basis. They did not receive the funds directly from the 2002 Settlement or the 1969 Settlement. The reasoning in paragraphs 104-130 does not apply to them. It would apply to them if the 2002 Settlement had distributed it to them directly but it did not: instead it paid it to Mr Bowring with no legal strings attached. Mr Bowring was not in law bound to pay the money on to his two cousins. Nor did he hold it on trusts which were essentially the same as the 1969 Trusts. “Tracing” does not apply because while the source of the money could be traced, it no longer was imprinted with a trust. Causation was broken.

[193]In any event, even if it could be said that the cousins received it indirectly from the 2002 Settlement (and we do not think it can), the 2002 Settlement was UK resident so s 87 does not apply in any event.

[194]Putting aside the position of the cousins, and considering the appeal in general, we note that it was not argued that:(a) On a purposive interpretation of the legislation and a realistic view of the facts (i.e. a Ramsay approach) s 90(5) would not operate to disapply s 90(1);

(b) That Mr Bowring was taxable under either s 97(5)(b) or (c)

(c) That there was a single settlement, being an “arrangement” (s 97(7) importing s 660G ICTA 1988) consisting of both settlements and the flip-flop machinery, which was a migrant settlement within s 89. And we express no view on (a) or (c) and no concluded view on (b) although as we have already said prima facie either or both 97(5)(b) or (c) might be applicable. Summary

[195]In conclusion we consider that Mr and Miss Bowring did receive all the distributions listed in paragraph 21 (and including the £400,000 passed on to the cousins) indirectly from the trustee of the 1969 Settlement and that the assessments should be upheld in their entirety. To echo Sir Stephen, to conclude otherwise would be shutting our eyes to the obvious.

[196]We dismiss the appeals.

[197]This document contains full findings of fact and reasons for the decision. Any party dissatisfied with this decision has a right to apply for permission to appeal against it pursuant to Rule 39 of the Tribunal Procedure (First-tier Tribunal) (Tax Chamber) Rules 2009. The application must be received by this Tribunal not later than 56 days after this decision is sent to that party. The parties are referred to “Guidance to accompany a Decision from the First-tier Tribunal (Tax Chamber)” which accompanies and forms part of this decision notice. BARBARA MOSEDALE TRIBUNAL JUDGE RELEASE DATE: 25 June 2013

Cited in 1 later judgment

Sarah Jane Turcan & Ors v The Commissioner For HMRC 2024