In Clive Bowring and Juliet Bowring v HMRC11, the Upper Tribunal (UT) has allowed the taxpayers’ appeal and concluded that a scheme, designed to reduce capital gains tax due on capital payments made by a trust, was effective.


Clive Bowring and his sister Juliet Bowring (the Appellants), appealed to the UT against a decision of the FTT. The FTT had dismissed their appeals against closure notices which had amended their self-assessment tax returns for 2002-03, to the effect that they were liable to capital gains tax of £849,644 and £317,417, respectively, as a result of additional gains under section 87, TCGA and supplemental charges under section 91 TCGA. All statutory references below are to TCGA.

The Appellants had been beneficiaries of an offshore discretionary settlement created by their father in 1969 (the 1969 trust) which, by 2001-02, had trust gains (within the meaning of section 87(2)) of some £3 million. By virtue of section 87(4), these gains would be treated as chargeable gains accruing to the beneficiaries of the 1969 trust who received distributions from the trustees and would give rise to a capital gains tax charge at a total rate of 64%.

In 2002, Clifford Chance (Mr Bowring’s solicitors) wrote to Mr Bowring commenting that it was likely that the trustee of the 1969 trust would make further distributions to the Appellants and suggested that planning known as “flip-flop mark II” be implemented before any such distributions were made.

In pursuance of this planning, another discretionary settlement was created in 2002, with the Appellants as beneficiaries (the 2002 trust). The assets of the 1969 trust were sold and the proceeds used to purchase £4 million worth of gilts. The trustee of the 1969 settlement borrowed £3.8 million on the security of the gilts, which was transferred to the 2002 trust.

Subsequently, distributions of £2.4 million were made to the Appellants by the trustees of the 2002 trust. In May 2002, the trustee of the 1969 trust sold the gilts and repaid the loan.

The Appellants argued that the source of the distributions was the 2002 trust alone, which meant that section 90(5)(a) operated to prevent section 90(1) from applying to the transfer, so that no tax liability was incurred on the distributions. HMRC argued that the capital payments had been made by the 1969 trust via the 2002 trust acting as an intermediary, and therefore a transfer of settled property did not occur within the meaning of section 90(1).

The parties agreed before the FTT that for the purposes of section 97(5), the distributions could not be regarded as received from both sets of trustees, as this would give rise to multiple charges to capital gains tax which Parliament could not have intended.

The FTT, in dismissing the Appellants’ appeals, held that section 97(5) should be interpreted widely, so that a capital payment could be regarded as received by a beneficiary from the trustees of one trust directly and from the trustees of another trust indirectly and that section 87(5) would prevent any multiple taxation which might otherwise arise.

The UT’s decision

The question for determination by the UT was whether the capital payments received by the Appellants should be treated as if made from the 1969 trust rather than from the 2002 trust, for the purposes of section 97(5).

The Appellants argued that it was not possible to regard the capital payments as other than made by and received from the 2002 trust. HMRC argued that, taking a realistic view of the  facts and applying the “signposts” referred to in Herman v HMRC12, the capital payments had been made by the 1969 trust via the 2002 trust acting as intermediary. Accordingly, “transfer … of … settled property” to the 2002 trust did not occur within the meaning of section 90(1) and therefore there would have been no transfer of trust gains between the two settlements, even if section 90 had not been ‘switched off’ by section 90(5)(a).

In the view of the UT, the FTT had erred in holding that section 97(5)(a) permitted the same capital payment to be treated as having been “received … from” the trustees of one settlement directly and from the trustees of another settlement indirectly. Section 87(5) had   to be construed as operating in respect of receipts from a single settlement in a single year  and could not provide a solution to the risk of double taxation which might arise if a single capital payment could be “received … from” more than one trust. The absence of any means of avoiding the risk of multiple taxation, and other anomalies and uncertainties to which the FTT’s construction of section 97(5)(a) would give rise, was a strong indication against the adoption of it. In addition, section 97(5)(a) was framed in terms which naturally appeared to envisage that a particular payment should be “received ... from” a single trust, either directly or indirectly. The natural meaning and effect of indirect receipt in section 97(5)(a) was that a beneficiary was still to be taken to have received a payment from a trust even if it was paid to him through an intermediary. It was conceivable that a separate trust could constitute such an intermediary, but the payment would not, in such circumstances, fall to be treated as “made by” and “received from” that intermediary.

HMRC’s approach to section 90 ignored the clear words and intended effect of the legislation. Section 90 referred to a straightforward legal concept, namely, a transfer of settled property.  Such a transfer took place if property ceased to be settled property of the transferor trust and became settled property of the transferee trust. Section 90 expressly acknowledged such a transfer by ensuring that relevant trust gains were also transferred, because it assumed it was from the trustees of the transferee trust that capital payments from the transferred property would be received. The UT said that no real exercise in interpretation or construction was required in order to ascertain whether such a transfer of settled property had occurred and in the instant case it clearly had. It was not possible to construe section 90 in such a way that the actual transfer was treated as not having occurred for tax purposes. Such a construction would reflect neither a realistic view of the facts or be consistent with the clear purpose of the legislation.

In the view of the UT, the trustee of the 1969 trust made an outright and unconditional transfer of its settled property to the 2002 trust. There was no agreement between the trustees of the two trusts, and the trustee of the 1969 trust had no say in what the trustees of the 2002 trust did with the transferred property. The UT concluded that it was not possible, taking a realistic view of the facts, to regard the 2002 trust as a mere intermediary in the sense that would be necessary if the distributions were to be treated as “received ... from” the 1969 trust indirectly. The capital distributions were clearly “received ... from” the 2002 trust and accordingly the appeal was allowed.


Although the arrangements had envisaged virtually all the transferred property being paid to the beneficiaries of the 2002 trust and the trustees of both trusts had knowingly implemented the scheme, this did not change the fact that the settled property in the 1969 trust had been transferred to the 2002 trust. Accordingly, when the 2002 trustees made the capital payments, they did so entirely in the exercise of their own discretion. Whilst HMRC no doubt considered the arrangements entered into as constituting unacceptable tax avoidance, the UT has confirmed that it is not always possible for HMRC to strain the facts in order to produce the outcome it desires. It would not be surprising if HMRC was to seek to pursue an appeal to the Court of Appeal.

The decision can be read here.