Barker v Baxendale Walker Solicitors (a firm) & Anor

This is a professional negligence case in relation to specialist tax advice relating to a tax avoidance scheme designed to avoid both Capital Gains Tax and Inheritance Tax. The scheme was based upon the establishment of an employee benefit trust (an “EBT”) offshore, the transfer of company shares by way of gift to the EBT and the creation of a sub-trust, the beneficiaries of which included members of the transferor’s family albeit that they were excluded from benefit until after the death of the transferor.

The issue is whether a reasonably careful practitioner with the degree of expertise claimed by the Respondents should have warned the Appellant, Mr Barker, before he transferred his shares in Team 121 Holdings Limited (the “Company”) to the trustees of an EBT that there was a significant risk that the EBT would fail to deliver the hoped for tax advantages because it did not exclude as beneficiaries after his death, his family members, being persons connected with him in his capacity as a “participator” in the Company. I will refer to such a warning as a “specific warning” and to the relevant risk as “the post-death exclusion construction” in relation to section 28(4) Inheritance Tax Act 1984 (“IHTA”).

More than a decade after the tax avoidance scheme had been established and the advice given, Her Majesty’s Revenue and Customs (“HMRC”) raised assessments on Mr Barker and challenged the validity of the EBT scheme on the basis of the post-death exclusion construction. Having been advised that the challenge would probably succeed, Mr Barker entered into a settlement with HMRC under which he paid approximately £11.29m by way of tax and interest. Mr Barker then took steps to seek to unravel the EBT scheme and sought damages in negligence against the Respondents.

In his judgment dated 23 March 2016 Roth J held that in the course of advising Mr Barker, the Respondents should have made clear that since the transfer of the shares to the EBT was a tax avoidance scheme there was a possibility that it would be challenged by HMRC, that if it were necessary to defend the arrangements in legal proceedings, there was a possibility that they would not be upheld and that in failing to do so they were in breach of their duty of care. However, he went on to find at [174] that such a “general health warning” would not have deterred Mr Barker from going ahead with the EBT scheme. The Judge accepted at [176], however, that had the specific warning been given, which he referred to as a “high level warning”, Mr Barker would not have proceeded with the EBT structure. Nevertheless, at [178] Roth J concluded that it was not a case in which any careful and competent solicitor of appropriate expertise would have given the specific warning. Mr Barker contends that such a specific warning should have been given and that the Respondents were in breach of their duty of care in failing to do so.

The First Respondent, Baxendale Walker Solicitors, is a firm of solicitors which specialised in tax advice, in particular, regarding tax planning and avoidance schemes (“BWS”). It was not represented before Roth J and has not been represented on this appeal. Apparently, its business ceased a considerable time ago. The Second Respondent, Mr Baxendale-Walker, was its only equity partner. It is common ground that both the case at first instance and the appeal stand or fall against both Respondents although only Mr Baxendale-Walker has been represented before the Court.

Mr Barker was the co-founder and majority shareholder of the Company. Mr Barker had set up the business in 1991 and it developed into a group of trading companies of which the Company was the parent. By 1998, the group had approximately 450 employees trading in several countries, although it remained UK based. At that stage, Mr Barker and his fellow shareholders decided to sell the Company. Since the group was considered to be worth £30-40 million, the tax consequences of such a sale were considerable and Mr Barker decided to seek tax planning advice to mitigate his exposure to Capital Gains Tax on the sale of his shares. He took advice from Deloitte & Touche (“Deloitte”) who prepared a report in June/July 1998 recommending three options for him, one of which was to set up a private unit trust scheme which would defer the payment of Capital Gains Tax until either Mr Barker was no longer tax resident in the UK or the assets had fallen into his estate.

In September 1998, Mr Barker met with Mr Nigel Hollinshead of Taxation Practical Services Ltd which had provided tax advice to the Company, its directors and employees for some years. It was Mr Hollinshead who suggested the possibility of an EBT and arranged an introduction to BWS on the basis that they specialised in this area. Mr Barker and a number of his fellow shareholders met with Mr Auden of BWS on 30 September 1998 and they were advised to adopt an EBT structure.

Although the precise terms of the legislation, and section 28(4) IHTA in particular, are central to the issues on this appeal and I shall turn to them in detail below, it is important to note at this stage that in general terms, an EBT is a trust for the benefit of employees of a company or body which attracts generous tax concessions. The trustees of the EBT must hold more than 50% of the shares in the company in question and the settled property must not be applied otherwise than for the benefit of employees of the company and their families and dependants and the class of beneficiaries must include all or most of the persons employed or holding office with the company.

Mr Barker met with Mr Baxendale-Walker at BWS’s offices in October 1998 and thereafter, Mr Barker was provided with a memorandum dated 16 October 1998. Under the heading “Summary”, at paragraph 3.1 it stated that Mr Barker’s objective was for the proceeds of sale of his shares to be enjoyed free of Capital Gains Tax and Inheritance Tax and at paragraph 3.6 stated that in the Respondents’ opinion the objective could be met by the use of “a type of Employee Benefits and Shares Trust (known as an “EBT”) as a tax efficient vehicle for providing the proceeds of sale of the Shares to the Taxpayer [Mr Barker] and his family and others.” It contained the recommendation to set up an EBT with offshore trustees and to execute a deed of gift transferring the beneficial interest in Mr Barker’s shares to the trustees of the EBT. It was expressly stated that as Mr Barker and his wife were controlling shareholders in the Company, the Deed establishing the EBT should provide that neither of them nor (during their lives) their children and remoter descendants should be entitled to receive outright distributions of income or capital from the EBT. It went on, however, to state that after the death of Mr Barker and his wife, their children and remoter issue could receive benefits outright from the EBT free both of Capital Gains Tax and Inheritance Tax.

On 6 October 1998, the Company created the EBT, the trustee of which was a Jersey company, Matheson Trust Co (Jersey) Ltd (“Matheson”). On 15 October 1998, Mr Barker executed a deed declaring that he held his shares in the Company on trust for the EBT. That deed contained a clause the purpose of which was to re-vest the shares in Mr Barker if HMRC determined that the EBT failed to meet the criteria of section 28 IHTA or section 239 TCGA. Thereafter, on 23 March 1999, Matheson declared a sub-trust, the beneficiaries of which were stated to be the widow, children and remoter issue of Mr Barker, his mother and sisters, living after his death. The Company was then sold on 28 June 1999 in a part shares part cash transaction to Logica Group Plc. The portion of the consideration for the sale referable to Mr Barker’s shares was notionally ring-fenced in the sub-trust.

After the transfer, a question arose as to whether Mr Barker should disclose the gift of his shares to the EBT in his tax return and Wragge & Co were instructed to consider Mr Barker’s compliance obligations as a result of the transfer and the EBT arrangement. On 29 July 1999, Wragge & Co wrote to Mr Barker setting out a number of lines of argument on the basis of which the transfer might be declared ineffective. These included the question of whether the conditions of section 28 IHTA had been met, but did not mention the post-death exclusion construction.

By 2002, Mr Barker had become disenchanted with Matheson, the Jersey trustee of the EBT, and was put in touch with Mr Andrew Brown, a chartered accountant who ran an independent tax consultancy and had experience in advising on offshore structures, including EBTs. By an email of 15 May 2002 from Mr Barker, Mr Brown was asked to advise on numerous topics including how Mr Barker might benefit from the EBT without violating its terms. Mr Brown responded in a letter of 10 June 2002 followed up by further advice in a letter of 24 June. He stated that he had no problems with the terms of the EBT itself and believed that it satisfied the conditions of the relevant tax legislation, including section 28 but went on to raise various other concerns. He advised again in March 2003 and the beginning of that substantive advice is set out at [75] of the Judgment. It refers specifically to Mr Barker’s children being within the class of persons allowed to benefit following Mr Barker’s death. However, the Judge went on to note that Mr Brown “said in evidence that when he wrote in those terms he was relying on the expertise of BWS and Mr Thornhill.” Mr Andrew Thornhill QC is an English barrister specialising in tax and was the co-author with Mr Baxendale-Walker of the book, “The Law and Taxation of Remuneration Trusts” first published in 1997.

Mr Brown suggested to Mr Barker the possibility of working with Mr Simon Bourge of Bourse Trust Co Ltd, (“Bourse”) a trust company based in Guernsey which provided offshore trust services, who through Bourse acted as a trustee for other EBTs. Mr Barker agreed to Mr Bourge being instructed and a meeting took place on 17 July 2002. Following the meeting, Mr Barker decided to appoint Bourse as trustee of the EBT in place of Matheson and agreed that Andrew Thornhill QC should be instructed to advise. The Judge records at [65] of the Judgment that Mr Bourge was eager to obtain Andrew Thornhill QC’s advice before he, Mr Bourge, became a trustee and had stated that one of the advantages of a new team of advisers was that they would examine “all of the assumptions, all of the trust documents, all of the previous advice, all of the correspondence from scratch with totally fresh eyes”. Mr Bourge did not, however, raise any questions about the ability of Mr Barker’s family to benefit from the EBT after his death.

Andrew Thornhill QC advised in September 2002. He was provided with copies of the EBT, the sub-trust deed and an extract from Wragge & Co.’s report. Amongst other things, he was asked to advise on the possible courses of action which might be open to enable Mr Barker to benefit from the funds within the EBT. Mr Thornhill conducted a review of the documentation and raised a number of concerns. However, he did not consider that the fact that Mr Barker’s family could benefit under the sub-trust after Mr Barker’s death presented a problem or rendered the EBT fiscally ineffective.

In addition, Mr Bourge also recommended that advice should be sought from Dr Raymond Ashton, a Guernsey Advocate. He was instructed in early November 2002 and was provided with the relevant documentation including the Wragge & Co report and a letter in response from BWS. He advised on the telephone that amongst other things, in his opinion, the gift to the EBT was “valid” and made no mention of the post-death exclusion construction. Following Dr Ashton’s advice, Mr Bourge was content that Bourse be appointed as trustee of the EBT and the sub-trust.

HMRC opened investigations in relation to the EBT arrangement in August 2005 and continued to investigate until 16 February 2007. The investigations were conducted by Mr Graham Hogg of the Capital Taxes Technical Group of HMRC in Edinburgh. He wrote to Mr Brown on 3 May 2006 and stated amongst other things that he had concluded that section 28 did not apply to the EBT. However, he analysed the EBT and sub-trust provisions in some detail and observed that pursuant to section 28(4), the terms of section 28(1) IHTA do not apply if the trusts permit any of the settled property to be applied at any time for the benefit of Mr Barker or any person who is connected with him.

On 19 March 2010, Mr Barker was informed by a letter from HMRC that further assessments for the years 1999/00, 2000/01 and 2001/02 would be raised. The reasoning for the assessments was explained in detail in a letter dated 19 April 2010 in which the post-death exclusion construction was referred to for the first time as one of the key bases of HMRC’s challenge to the EBT arrangement.

Mr Barker agreed with Mr Brown that Mr Thornhill QC should be instructed to advise upon HMRC’s arguments and in a written opinion of 5 July 2010, amongst other things, Mr Thornhill expressly disagreed with HMRC’s interpretation of section 28(4). Eventually on 16 February 2011, Mr Barker issued a formal Notice of Appeal against HMRC’s assessments and Farrer & Co (“Farrers”) were instructed to act as Mr Barker’s solicitors on the appeal. On Farrers’ recommendation, in July 2011, Mark Studer of counsel was instructed in relation to the trust law aspects of the case. In June 2012, Farrers sent further instructions to Mr Studer asking him to advise about various matters including the effect of the possible constructions of section 28(4) on the interpretation of the trust documents. He advised that there were “strong arguments” in favour of the post-death exclusion construction put forward by HMRC. Thereafter, Mr Thornhill QC accepted that Mr Studer’s view was “a strong one” at a meeting on 23 July 2012.

Following advice from his solicitors and accountant, on 17 April 2013, Mr Barker settled with HMRC for £11,296,477.24 and in the same month, issued a claim in negligence against Mr Baxendale-Walker and his firm. Proceedings were then commenced to reverse the transfer of Mr Barker’s shares to the EBT.

Meanwhile, on 1 October 2010, Bourse had been replaced as trustee of the EBT and sub-trust by Confiance Ltd, a Guernsey company (“Confiance”). In relation to the proposed reversal of the transfer of Mr Barker’s shares in the Company, Confiance as trustee instructed David Goy QC and Nicholas le Poidevin QC. In a joint opinion dated 3 April 2014, they concluded that on balance, HMRC’s view of the interpretation of section 28 was correct.

The Judge decided that the post-death exclusion construction was “not obvious or likely” and “very doubtful” although it was “arguable”: see the Judgment at [165], [156] and [177] respectively. He went on to state at [178] that he was “strongly inclined to the view” that Mr Baxendale-Walker’s construction of section 28(4) IHTA was correct. He approached the issue of whether a specific warning was necessary in the following way:

At this point, it is necessary to set out the relevant statutory provisions in detail. The general requirements for a valid EBT are set out in section 86 IHTA and the requirements to be met for IHT exemption for a gift to an EBT are set out in section 28 IHTA. The provision granting CGT holdover relief is set out in section 239(1) TCGA 1992. Hold-over relief is available where the requirements for IHT exemption are also met.

Sub-sections 86(1) and (3) IHTA are as follows:

Section 28 provides materially as follows:

“Participator” is defined in section 102(1) IHTA and section 417 Income and Corporation Taxes Act 1988. Mr Barker fell within that definition. Whether someone is “connected” with another person is defined in section 270 IHTA and section 286 TCGA. Mr Barker’s wife, children and remoter issue fall within the definition and are connected with him.

As I have already mentioned, Mr Barker’s ground of appeal is that Roth J was wrong to conclude that the Respondents did not act in breach of their duty by failing to give the specific warning. In particular, it is said that he failed to conclude that there was, at the very least, a significant risk that the Respondents’ construction was wrong and that “is” in section 28(4)(d) IHTA refers to the date at which assets are transferred to the trust rather than the date on which they are applied for the benefit of the persons in question, because: he failed to give proper weight to the fact that the construction which he preferred exposed the relief afforded by the section to tax avoidance abuse when passing wealth from one generation to another; he failed to give proper meaning to the words in section 28(4)(c) and required that paragraph to be re-written to make sense; and failed to take account of a serious ambiguity which arose on the Respondents’ construction as to the date at which the property is to be treated as applied for the benefit of the person for the purposes of the section. Further, it is said that he failed to give “at any time” its proper weight and meaning, failed to give a consistent meaning to the word “is” in paragraphs (a) – (d) of section 28(4) and placed too much weight both on the admitted lacuna which arises on the construction put forward on Mr Barker’s behalf and upon HMRC’s GAAR guidance as an aid to construction.

It is also said that the Judge wrongly took into account that other advisers failed to point out the existence of the post-death exclusion construction; and if he was entitled to take such failure into account, he should have concluded that it was outweighed by the fact that HMRC relied on the post-death exclusion construction and once they had done so all those who advised concluded that the construction was a strong one.

It is common ground that the relevant standard of care in relation to whether the specific warning should have been given is that of the reasonably competent practitioner in the field in question with the expertise claimed by the Respondents. It is also accepted that the question is fact sensitive and that the likely construction of the provision in question is highly relevant to the evaluation of whether a warning should have been given. I will turn first, therefore, to the question of construction.

In summary, Mr Furness on behalf of Mr Barker says that it is clear that the requirements of section 28(1)(a) have to be met at the date of transfer of value and that it is natural to expect that the criteria in sub-section (4) fall to be evaluated at the same time. In this regard, he points to sections 28(4)(a), (b) and (d) which each prohibit the trust from benefitting a person who “is” within the prohibited category. Mr Furness says that the use of the present tense “is” renders it natural to determine whether the test for prohibition is met at the date of the transfer and that such an approach gives full effect to the requirement that the prohibition must apply at “any time”. The section requires one to look at what the trust permits at the date of the transfer of value and to identify the potential beneficiaries by class at that stage. Accordingly, he says that if a person is a participator or a person connected with a participator at the date of the transfer, he must not benefit at any time thereafter and that those who become connected in the future also fall into the class. The fact that they may not be “connected persons” in the future is irrelevant. By contrast, he says that if the Respondents’ construction is right and one should look at the date when the property is applied, the phrase, “at any time” is given no weight and is, in fact, unnecessary.

Miss Campbell who made the tax submissions on behalf of Mr Baxendale-Walker accepted that the question of whether section 28 is satisfied must be considered at the outset when the transfer of value is made, albeit that she says the prohibitions in section 28(4)(a) - (d) need only be avoided at the date that the person in question has a benefit applied to him or her from the trust. In this regard, she relies upon the use of “is” in (a) and (d) which she says should be contrasted with “has been” in (c) and also points to the fact the phrase “at any time” is not within paragraph (a) itself. She submits, therefore, that someone who is initially in a prohibited category may nevertheless benefit at a later date, if they have ceased to be within that category. In other words, despite the fact that Mr Barker’s family would fall within section 28(4)(d) at the date on which the shares were transferred to the EBT, they would no longer be “connected” with a “participator” (Mr Barker) after his death and therefore, could benefit at that stage without falling within the prohibition.

Mr Furness submits that such a construction of “is” is untenable because of the wording of section 28(4)(c) IHTA which refers both to a person who has been a participator at any time during the ten years before the transfer and a person who becomes a participator at any time after the transfer. Neither may benefit. He accepts therefore, that “has been” in (c) should be construed to mean “who becomes or has been” in order to make sense of the reference in that paragraph both to the ten years before the transfer of value and any time after it. Mr Furness says therefore, that there is no room for an interpretation which, for example, permits a person to benefit when they are no longer a participator. They are excluded even if they ceased to be a participator even before the trust was created.

Mr Furness also points out that if Miss Campbell is correct, it is necessary to interpret section 28(4)(c) as if the following underlined phrases had been added:

Mr Furness also says that paragraphs (a) – (c) create a waterfall effect and that each expands upon the one before it. He accepts that the construction which he espouses may create a lacuna in the meaning of section 28(4)(d) because it would not naturally include those who became connected with a participator within paragraph (a), (b) or (c) after the transfer of value. For example, those who became spouses of a participator after the transfer of value would not be caught. He says that this is cured by construing “is” in paragraph (d) to include “or becomes”, or alternatively to construe “is” in a way that is “ambulatory” in the sense that it requires the test to be met at all times during the life of the trust and necessarily therefore, refers both to the present and to a future state of affairs.

Reliance was placed upon O’ Rourke v Binks [1992] STC 703 in which as part of the process of interpretation, words were read into a tax provision in order to avoid an anomaly. Scott LJ held at 710c that: “it would be a very rare case in which effect could not be given by a permissible process of interpretation to the apparent intention of the legislature. At all events, in the present case, I cannot accept that there is any reason why the limitation to which I have referred should not be implied, thereby giving effect to the presumed, the apparent, intention of the legislature.” He had already observed at 708e that the literal construction of the provision in question led to an anomaly which “merits being described as absurd.” Miss Campbell relies on the lacuna and says that the implications proposed by Mr Furness are impermissible.

Mr Furness also emphasises that section 28(4) disapplies sub-section (1) if the trusts permit the application of any of the settled property to any of the prohibited classes “at any time”. He submits that that phrase must be given effect and that by its use, coupled with the reference in parenthesis to the period referred to in section 86(1) “or later”, the draftsman makes it explicit that the prohibition extends to the entirety of the lifetime of the trust. He says that this avoids the very abuse of which HMRC complained in this case namely that a participator could make a transfer of value to an EBT on his death bed in the knowledge that his family would be able to benefit from the generous provisions of section 28 and 86 very shortly thereafter, and enjoy perhaps a substantial part of his estate tax free.

Miss Campbell relied upon section 28(1)(b) itself as the express means to avoid the use of the generous exemption in section 28 for tax avoidance. She pointed out that a settlement to which the exemptions apply must benefit all or most of the persons employed by or holding office with the company. She says, therefore, that that provision is a brake on the use of the exemption for dynastic tax planning. She also submitted that the policy reason for excluding connected persons in section 28(4)(d) is because the participator himself may benefit indirectly if they were included. However, that cannot be the case after the participator’s death and therefore, even if the Respondents’ construction is correct, the policy is fulfilled. In this regard, she referred to Inland Revenue Commissioners v Eversden & Anr (executors of Greenstock, deceased) [2003] EWCA Civ 668 at [22] for the proposition that it is beyond the court’s powers of interpretation to imply words into a tax statute in a way which would distort the language, merely to prevent what would otherwise appear to be tax avoidance.

Lastly, Miss Campbell referred us to HMRC’s GAAR Guidance, approved by an Advisory Panel with effect from 15 April 2013 to which the judge referred at [167] of his Judgment. At paragraph D29.5.1 reference is made to section 28 and it is stated that “the intention of the legislation is to provide an incentive to diversify share ownership or at least the benefits of share ownership among a wider group of people who are then incentivised to run the company profitably.” The following example and critique then follows:

The Judge dealt with construction of section 28(4) at [155] to [165] of the Judgment and Miss Campbell supports his reasoning. Mr Furness on the other hand says that the Judge’s conclusion is wrong and makes a number of detailed submissions. First, he says that the Judge was wrong at [158] to conclude that if Mr Barker’s construction were correct, it would be hard to see what paragraph (a) would add to (c). As I have already mentioned, he says that the provisions are naturally read as a waterfall and that furthermore, the Judge was wrong to conclude that the Respondents’ construction creates a clear distinction between (a) and (c) because in order to do so, it is necessary to do considerable violence to the wording by substantial insertions.

Secondly, in relation to the Judge’s reasoning at [159] to [161], Mr Furness submits that the Judge placed too great a weight upon the admitted lacuna in (d) which arises upon Mr Barker’s construction which can be cured by implication and that the problem is less serious than those which arise from the Respondents’ construction. Furthermore, he does not accept that the Respondents’ construction gives rise to no illogicality and requires no re-writing. Mr Furness says that it leaves the section wide open to abuse, requires the re-writing of (c) and carries with it the ambiguity to which he referred.

Thirdly, he takes issue with the Judge’s reasoning in [162] that “on the claimant's construction, since para (a) then refers to someone who is a participator at the time of the transfer of value, and para (c) on any view covers someone who was a participator in the 10 years before the transfer of value, although para (c) clearly also covers someone who has subsequently become a participator, I consider that the words “has been” do not readily include someone who becomes and remains a participator”. Mr Furness says that the Judge’s reasoning is fallacious and that (a) must be applied first and the fact that a participator may have been so before the date of the transfer of value is irrelevant. (c) is concerned with those who are not within (a).

Finally, Mr Furness says that the Judge was wrong at [163] to conclude that the phrase “at any time” does not provide much assistance because it is of great significance in the sub-section.

It seems to me that contrary to the Judge’s analysis, HMRC’s construction of section 28(4) is very likely to be correct. First, I agree with Mr Furness that it is clear that the requirements of section 28(1)(a) must be met at the date of the transfer of value and that it is natural therefore, that the exception to the exemption found in section 28(4) should be evaluated at the same time. I consider it to be entirely artificial to say that although that evaluation of whether “the trusts permit any of the settled property to be applied at any time . .” for the benefit of the persons mentioned in (a) to (d) should be undertaken at the date of the transfer of value, nevertheless, the question of whether the criteria which need to be avoided are met must be judged by reference to the date at which a benefit is applied for a beneficiary. It seems to me that such a construction is unnatural and requires a good deal of re-writing/implication of words. Furthermore, it also creates an ambiguity in relation to the meaning of “applied” in relation to a benefit. Is that a reference to the time at which the beneficiary becomes beneficially entitled to settled property or when it is appropriated to him by the trustees? It seems to me that if that had been the intention of the legislature, the entirety of section 28(4) would have been drafted differently or, at the very least, the term “application of settled property” would have been defined.

Furthermore, in my judgment, such a construction fails to give proper weight to the phrase “at any time” in the opening part of section 28(4) particularly when that phrase is read together with the phrase in parenthesis, which makes clear that the draftsman intended the prohibition to last for the entirety of the life of the trust and not just for any period during which it satisfied section 86(1). Furthermore, it seems to me that given the use of the word “is” in paragraphs (a), (b) and (d) of section 28(4), it is natural to test whether any of the prohibitions are met at the date of the transfer of value and that such a construction gives full effect to the phrase “at any time.” It follows that I disagree with Roth J at [163] of his judgment when he said that he derived little assistance from the words “at any time”.

I also disagree that the use of “is” in paragraphs (a), (b) and (d) when contrasted with “has been” in paragraph (c) leads to the conclusion that the question of whether a person falls within a prohibited category should only be judged at the date on which settled property is applied for his benefit. Nor do I consider that it creates the difficulties which the Judge envisaged at [163] of his Judgment. It seems to me that if the Respondents’ construction were right it would be necessary to interpret paragraph (c) in the rather unnatural way to which Mr Furness referred and which I have set out at paragraph 32 above and for which I can see no justification.

It also seems to me that each of paragraphs (a) – (c) of section 28(4) are mutually exclusive and that therefore, the construction which troubled the Judge at [158] of his judgment does not arise. Section 28(4)(c) begins with the phrase “any other person who has been a participator” (emphasis added) and therefore the natural meaning of that paragraph is that it does not override or subsume that which has come before. Therefore, it does not render (a) otiose as the Judge concluded. It relates to “any other person” in addition to those already mentioned in (a) and (b). It seems to me that (c) read as a whole and in the context of section 28(4), refers to a person who is not within (a) or (b) and therefore, is not a participator at the very date of the transfer of value but either was a participator at any time in the ten year period before the transfer or becomes a participator at any time thereafter. As Mr Furness points out, if Miss Campbell’s construction were correct, paragraph (a) would have no purpose at all.

Furthermore, I do not share the Judge’s concerns expressed at [159] and [160] of his Judgment about the need, on Mr Barker’s construction, to imply “or becomes” into paragraph (d) or to construe “is” in (d) in a way which means that it has to apply at all times throughout the life of the trust. It seems to me that if one reads (d) in the light of section 28(4) as a whole and paragraph (c) in particular, it is a perfectly natural construction or implication to apply. Paragraph (d) is intended to apply in relation to any person whether in (a), (b) or (c). Therefore, it needs to be read in a way which covers all of the temporal circumstances covered in those earlier paragraphs. I can see no reason why “is” should not be read in an “ambulatory” fashion or for that matter, that “or becomes” should not be implied. This situation is a long way from the circumstances under consideration in IRC v Eversden.

Lastly, as Mr Furness points out, Mr Barker’s preferred construction also avoids the abuse to which HMRC referred in their letter of 19 April 2010 of the death bed transfer of value. It also avoids attributing to Parliament the implausible intention that an employee benefit trust could be used for dynastic estate planning and enable the family of the owner of a major shareholding in a company to benefit from the proceeds of sale of that holding entirely tax free after the owner’s death. It seems to me that Miss Campbell’s argument that section 28(1)(b) is itself a brake on the use of section 28 for dynastic planning falls foul of the facts of this case. Mr Barker satisfied that provision but still hoped to use the section for that very purpose. I view her submissions in relation to the purpose behind the exclusion of connected persons in the same way. She says that the purpose of their exclusion is to prevent the participator from indirect benefit from the EBT. However, it seems to me that it must be of benefit to the participator to have the potential to pass on some of his estate to his family entirely tax free, as Mr Barker’s intentions prove.

In my judgment, therefore, Mr Baxendale-Walker’s construction of section 28(4) is very unlikely to be correct. It requires a great deal of implication, far beyond that required by Mr Furness’ approach, gives little or no weight to “at any time” and creates an ambiguity as to the meaning of “applied”.

As I have already mentioned, it is accepted that the relevant standard of care in relation to whether the specific warning should have been given is that of the reasonably competent practitioner in the field in question with the expertise claimed by the Respondents. However, the parties differ about the application of the second part of the approach set out in Bolam v Friern Barnet Hospital Management Committee [1957] 1 WLR 582 at 587, namely that the standard of care of the reasonably competent practitioner should be assessed by reference to whether the Respondents were acting in accordance with the practice of competent respected professional opinion. Mr Furness QC on behalf of Mr Barker says that there is no need to take account of an actual body of professional opinion when determining whether the standard of care was breached. He says that it is for the court to determine whether the specific warning should have been given on the basis of what a reasonably competent lawyer with the Respondents’ expertise would have done in the circumstances. He says, therefore, that Roth J erred in taking into account the views of the “series of experienced tax specialists” to whom he referred in [178] of his Judgment. He submits that in any event, the views which were taken into account could not be characterised as a body of competent respected opinion.

Mr Furness referred us to a series of cases in which the court decided whether a solicitor had been negligent in failing to set out the risk of his view of a particular provision being wrong without reference to the evidence of a body of respected professional opinion. In some cases, as here, it was not alleged that the solicitor’s advice as to the construction of the provision itself had been negligent. Mr Furness relied upon Queen Elizabeth’s Grammar School Blackburn Ltd & Anor v Banks Wilson (A Firm) [2001] EWCA Civ 1360, [2002] PNLR 14, for both those propositions. That was a case which was concerned with advice given about the meaning of a restrictive covenant which had been drafted by the solicitors who had acted for the purchaser of the site in question. After building work had commenced, the vendor in whose favour the covenant had been granted went on site and saw the plans and stated that he considered that in breach of the covenant the proposed building would be too high. Advice was given to the purchaser at that stage that as long as the new building did not exceed the height of the chimney pots of the previous buildings the covenant would not be infringed. Ultimately, the building was re-designed and built with a lower roofline and the purchaser alleged that the solicitors’ advice had been negligent. The Judge at first instance found that the solicitors were neither negligent in their original advice as to the meaning of the covenant nor in their advice after work had begun once the issue had been raised by the vendor/covenantee.

The question for the Court of Appeal was whether there was “real scope for dispute” about the meaning of the restrictive covenant which the solicitor had drafted. In relation to the advice given in October 1994, after the point had been taken on site, the respondent accepted that if there were such scope, the solicitor should have pointed out the risks to his client. Arden LJ approached the matter in the following way:

The matter was considered more recently in Balogun v Boyes Sutton and Perry [2017] EWCA Civ 75; [2017] PNLR 20. That was a case in which an under-lessee contended that solicitors had failed to provide him, as instructed, with adequate advice as to the permission he needed from the superior landlord, in order to use a ventilation shaft which was essential to the operation of his restaurant. He contended that the under-lease did not correspond with the head lease and as a result he did not enjoy a right to connect to the ventilation shaft. At first instance, the judge found that no such instructions had been given. On appeal the claim was also put on the alternative basis that even if express instructions had not been given, the solicitors should have advised their client to seek agreement to amend the underlease as the wording of the draft created a risk of argument. Lloyd Jones LJ (as he then was) with whom King and Gloster LJJ agreed, came to the same conclusion as the judge at first instance, namely that on a proper construction the under-lease did create a right to connect to and use the ventilation shaft. Nevertheless, he went on:

Mr Furness also relies upon Herrmann v Withers LLP [2012] EWHC 1492 (Ch); [2012] PNLR 28 and Levicom International Holdings BV v Linklaters [2010] EWCA Civ 494; [2010] PNLR 29. In the Herrmann case, Newey J, as he then was, decided that Withers were not open to criticism for taking the view they did about the construction of the Kensington Improvement Act 1851 which gave occupiers of houses on a particular square the right of access to communal gardens. However, he decided that they were at fault in failing to appreciate that there was a substantial risk that their construction might not be right and that a reasonable solicitor would have realised that the point was open to question. See [68] and [71] of the judgment. He made no reference to a body of professional opinion in relation to whether the Herrmanns should have been advised of the risk. In the Levicom case, Stanley Burnton LJ with whom Lloyd and Jacob LJJ agreed, held that although it would be wrong to decide that the defendant solicitors were negligent to advise that the construction of a particular clause in a share sale agreement was correct, they had failed to take account of various factors and could not have sensibly advised that the breach of the particular clause was “clear”. He decided that they were negligent in doing so and that it was “particularly relevant to give a balanced view in the context of potential arbitration proceedings . . ”. See [249] of the judgment.

Mr Furness submits, therefore, that the Judge was wrong to place any reliance upon the various professionals who failed to pick up on the post-death exclusion construction point before it was raised for the first time in 2010 and that, in any event, the opinion which the Judge relied upon was only that of an unrepresentative group of individuals rather than expert evidence as to a body of professional opinion, there having been no expert evidence at the trial.

He also says that, if, in this case, the court considers that the construction placed upon section 28 IHTA by HMRC is likely to be correct, then the Respondents will have been negligent in not having given the specific warning. However, he says that even if the Respondents’ construction were correct, the risk of it being wrong was sufficiently great that in all the circumstances of this case, a warning should still have been given. He says that there was quite clearly “real scope for a dispute” and a “significant risk” that the EBT would fail to achieve its intended fiscal purpose. He submits, therefore, that the Judge placed the bar too high when he stated at [178] of the Judgment that perhaps the arguments for and against a construction would have to be “finely balanced” before solicitors whose interpretation is likely to be correct would be in breach of duty in failing to warn the client that their interpretation may be wrong.

Mr Seitler QC submits that reasonable competence for the purposes of the approach in the Bolam case must be assessed by reference to whether the Respondents were acting in accordance with a reasonable body of professional opinion in the field and that therefore, Roth J was right to take into account the views of the other professionals who considered the EBT in the late 1990s and early in the new century and failed to draw attention to the post-death exclusion construction. Mr Seitler includes Roth J himself in the list of those who took the Respondents’ view on construction and says that the Judge’s conclusion should also be taken into account when determining whether the Bolam test is satisfied. Furthermore, he says that the correct question to ask is whether no reasonably competent practitioner could have failed to give the specific warning.

He says that one should only depart from such an approach in very limited circumstances and that both the recent cases of Montgomery v Lanarkshire Health Board (General Medical Council intervening) [2015] UKSC 11 and O’Hare v Coutts & Co [2016] EWHC 2224 (QB) which were raised by Mr Furness in written argument, can be explained in that way. The Montgomery case concerned a consultant’s omission to advise his patient about a particular known risk of a natural delivery of her baby. Having considered the development of the doctor patient relationship and the changes in the way in which health providers and recipients of services are viewed, at [74] and [75] Lords Kerr and Reid JJSC, with whom Lords Neuberger, Clarke, Wilson and Hodge JJSC agreed went on as follows:

Mr Seitler submits that a Montgomery style standard should not be applied by default in this case. Although the extent to which a doctor may be inclined to discuss risks with his patient is not a matter of medical learning or expertise the same is not true in relation to the risks surrounding the construction of section 28(4) which is clearly a question of legal learning. He submits that the O’Hare case, equally, has limited application. In that case, Kerr J was considering investment advice and declined to apply the Bolam approach to proper dialogue and communications between a financial adviser and his client to ensure the client understands the advice and the risks attendant on a recommended investment and applied the reasoning in the Montgomery case instead.

In this case, it is not alleged that the Respondents were negligent to take the view of the construction of section 28(4) that they appear to have done. It is accepted that a reasonably competent solicitor with the Respondents’ expertise could have done so. The question is whether in the light of all the circumstances no reasonably competent solicitor in the position of the Respondents would have failed to give the specific warning that there was a significant risk that the EBT arrangement would fail to be tax effective because of the post-death exclusion construction. This is the way in which Mr Barker’s ground of appeal is founded and it seems to me that it follows that the usual test applies when determining the scope of the solicitor’s duty and whether it has been breached.

It is important to appreciate that the court is considering what advice ought to have been given by the reasonably competent practitioner in the particular factual circumstances at the time. Of course, the advice which a reasonably competent solicitor would give in the circumstances turns substantially upon the view that he could take of the provision on which it turns. It is also dependent upon whether contrary arguments as to construction are of sufficient significance to require specific mention when taken with the degree of risk inherent in the circumstances and the importance in those circumstances of a balanced view of the provision. As Salmon LJ noted in Dixie v Parsons the question turns in part upon the likelihood or otherwise of a dispute.

It seems to me that the following principles are likely to apply:

The question of whether a solicitor is in breach of a duty to explain the risk that a court may come to a different interpretation from that which he advises is correct is highly fact sensitive: (Queen Elizabeth, Hermann, Balogun and Levicom cases);

If the construction of the provision is clear, it is very likely that whatever the circumstances, the threshold of “significant risk” will not be met and it will not be necessary to caveat the advice given and explain the risks involved;

However, depending on the circumstances, it is perfectly possible to be correct about the construction of a provision or, at least, not negligent in that regard, but nevertheless to be under a duty to point out the risks involved and to have been negligent in not having done so (Levicom and Balogun cases);

It is more likely that there will be a duty to point out the risks, or to put the matter another way, that a reasonably competent solicitor would not fail to point them out when advising, if litigation is already on foot or the point has already been taken, although this need not necessarily be the case (the Queen Elizabeth case to be compared with Balogun); and

The issue is not one of percentages or whether opposing possible constructions are “finely balanced” but is more nuanced.

The issue of whether there was a “significant risk” in this case turns to a considerable extent upon the likely construction of section 28(4) and upon the factual circumstances including the very nature of the EBT arrangement itself and its purpose, the amount at stake and the risk of challenge. This conclusion is consistent with the approach taken in the Withers, Levicom, Queen Elizabeth and Balogun cases. It seems to me therefore, that the question of whether there was a significant risk that the EBT would fail to deliver the promised tax advantages because of the post-death exclusion construction, is one of law or at least, mixed fact and law to be determined by the court applying the standard of the reasonably competent solicitor.

I should add that it seems to me that the position is different from the situation under consideration in the Montgomery case. Much of the discussion of the Supreme Court is focussed on the relationship between doctor and patient and the changing social mores. The duty under consideration was a duty to take “reasonable care to ensure that the patient was aware of any material risks involved in any recommended treatment” and was considered separately from the duty to diagnose and treat the patient which were questions of medical knowledge and expertise. The question with which the Supreme Court had to grapple was whether there was a separate duty to inform or warn. It is not surprising therefore, that a different test, one of materiality, was applied and that the question of whether a warning/information should have been given to the patient was not considered to be a matter of medical learning and accordingly, was not governed by the approach in the Bolam case.

In this case, legal advice was the very service which was being provided and which was being relied upon. There can be no separation between the advice and any appropriate caveats as to risk. They are one and the same. The lawyer as part of the legal advice he is providing, must evaluate the legal position and determine whether in all of the circumstances, he should advise his client that there is a significant risk that the view he has taken about the substantive matter in question may be wrong: see the Queen Elizabeth case per Arden LJ at [29] and the Balogun case at [36] and [38]. As I have already mentioned, it seems to me that that is a question of law and legal expertise and not a policy question. It is not necessary for me to decide whether the Bolam test is applicable in the case of financial advisers in circumstances such as those considered by Kerr J in the O’Hare decision.

When determining whether a reasonably competent adviser would have advised that there was a significant risk that a contrary view would be taken in relation to section 28(4) and that the post-death exclusion construction might well be correct, the relevant facts included the fact that this was a very aggressive tax avoidance scheme which was marketed to Mr Barker on the very basis that his family would be able to benefit from the property within the EBT at the date of his death free of Capital Gains Tax and Inheritance Tax, an outcome which might appear on the face of it to be too good to be true. It was for that reason that the sub-trust was established at the outset and section 28(4) and paragraph (d) in particular, were the focus of the drafting and ought to have been at the centre of the advice. It is also relevant that the potential charge to tax was very large and the Respondents’ fee was in the region of £2.4m.

In my judgment, Mr Seitler’s point that there is no evidence that Mr Baxendale-Walker had any knowledge of the possibility that Mr Barker might enter into another arrangement and therefore, it was an all or nothing situation, holds no water. It can hardly be relevant that the vendor of the tax avoidance product has no knowledge of whether if warned of a substantial risk which goes to the heart of the efficacy of the product, the purchaser would take a different course or not.

I also consider that given the amounts at stake and the very nature of the aggressive tax avoidance scheme there was always a likelihood of a dispute which should have been taken into consideration. Although just as in the Balogan case, the point had not been taken already (as it had been in the Queen Elizabeth case) it would have been obvious to any reasonably competent solicitor practising in this area that there was a real risk that HMRC would take the post-death exclusion construction point at some stage and if necessary, would pursue it through the tribunal and court system, especially as the EBT arrangement was founded on the ability of Mr Barker’s family to benefit after his death which was its purpose and there was a considerable amount of tax at stake. To put the matter in the words of Sedley LJ in the Queen Elizabeth case, the post-death exclusion construction, which was at the very heart of the EBT arrangement, was likely to cause “a lot of trouble.”

As will already be apparent, I also consider that the position taken by other advisers at the time is irrelevant. The individuals, Mr Thornhill, Dr Ashton, Mr Bourge, Mr Brown and Mr Hogg who considered the EBT for different purposes and at different times were at best, an unrepresentative group. Even if it had been relevant to consider the standards applied by a body of expert professional opinion, they were not such a body. It is common ground that there was no expert evidence before the court. Furthermore, as Mr Furness pointed out, Mr Thornhill who had collaborated closely with Mr Baxendale-Walker in the past, changed his mind and back-tracked in 2012 accepting that there were strong arguments in favour of the post-death exclusion construction. Dr Ashton, a Guernsey lawyer, looked at the EBT for different purposes and advised as to the validity of the gift to the EBT. Mr Bourge was the representative of a Guernsey trust company who was eager that the advice of Mr Thornhill be taken before his company took on the trusteeship of the EBT and does not appear to have advised expressly himself at all. Mr Hogg of HMRC was not a lawyer and therefore, could not form part of a body of legal opinion for the purposes of the Bolam test. In addition, the Judge recorded that Mr Brown, an accountant had relied upon the advice of Mr Thornhill and BWS.

Further, to give their conclusions any weight or to derive comfort from them, as the Judge did at [178] of his Judgment is to ignore those whose opinions went the other way, including Mr Le Poidevin, Mr Goy, Mr Studer and Mr Thornhill at the second time of asking. It is also to ignore the fact that Mr Barker was advised to settle with HMRC and did so. I should add that despite Mr Seitler’s submission to the contrary, I do not consider the fact that the Judge came to a different conclusion on the interpretation of section 28(4) is relevant for the purposes of the Bolam test. If that were the case, the appeal process would be weighted in the Respondents’ favour.

Equally, it seems to me that the content of the GAAR guidance to which I referred at [45] and which the Judge stated at [167] had “buttressed” his conclusion is of little assistance when determining whether there was a “substantial risk” in all the circumstances.

It follows that in my judgment, Roth J was wrong about the construction of section 28(4) and his view of the construction coloured his approach to the risks involved and whether it was appropriate therefore, to give the specific warning. In my judgment, there was a significant risk that the EBT arrangement would not work as a result of the post-death exclusion construction which was centrally important to its structure and the likelihood that the promised tax advantages would be delivered. In all those circumstances, despite the fact that it is not alleged that the Respondents’ view of the construction of section 28 was itself negligent, they should have given the specific warning. There was a significant risk that their advice was wrong and in all the circumstances, a reasonably competent solicitor would have gone beyond his own view and set out the risks.

I also consider that in the light of the decision in the Balogun case, Roth J was wrong to conclude that where the solicitor’s interpretation of the particular provision is likely to be correct it is difficult to see that they would also be in breach of a duty to warn unless the arguments were finely balanced. Although the question of whether there is real scope for dispute or a significant risk turns in substantial part upon the strength of the arguments for and against the particular construction, it also depends upon all of the relevant circumstances. Hence the decision in Balogun itself.

For all of the reasons set out above, I would allow the appeal.

I agree. Since, however, we are differing from a very experienced judge, whose views are always entitled to great respect; I would like to explain briefly in my own words why I too have come to the conclusion that HMRC’s construction of section 28(4) is probably correct.

The evident purpose of IHTA section 28 is to encourage the transfer of shares or securities in a company into an EBT by providing an exemption for an individual shareholder who makes such a transfer, provided that certain conditions are satisfied. If the conditions for exemption under section 28 are satisfied, it then follows automatically that holdover relief from CGT will also be available by virtue of section 239(1)(b) and (2) of CGTA 1992.

This generous fiscal treatment would, however, be open to obvious abuse if it were possible for an individual shareholder to obtain exemption from IHT under section 28, and to avoid an immediate charge to CGT based on the current market value of the shares transferred, but the trusts of the EBT were such that it could be used as a vehicle to provide future capital benefits for himself or members of his family or other chosen objects of his bounty. The possibility of such abuse is all the more apparent because the definition of an EBT, in section 86(1) of IHTA, refers to trusts for the exclusive benefit of employees (together with their spouses, relations and dependents) “either indefinitely or until the end of a period (whether defined by a date or in some other way)” (my emphasis). An EBT may accordingly be established for a limited period, of no specified minimum duration. There is nothing in the definition itself to prevent the settlement in question from containing trusts in favour of a wider class or classes of beneficiaries, including the settlor himself, his family, or persons with no prior connection with the company, once the limited period has expired.

In broad terms, it is the possibility of this kind of misuse which section 28(4) must have been designed to counter. Prevention of the avoidance of capital taxation is the whole raison d’être of the subsection: no other purpose for it has been suggested. The subsection operates by disapplying the provisions of subsection (1), with the result that the transfer of value made by the individual shareholder no longer qualifies for the exemption conferred by the subsection, and the CGT holdover relief therefore falls away as well.

Since subsection (4) operates by disapplying subsection (1), and since the question whether a transfer of value is exempt from IHT necessarily has to be judged as at the time of the transfer, it follows that the primary focus of subsection (4) must also be on circumstances as they exist at that time. This point is reinforced by the use of the present tense in the opening words of both subsections:

In answering that question, regard must obviously be had to the various ways in which it may be or become possible for the trustees to apply the trust property, or any part of it, consistently with the terms of the settlement and the requirements of trust law. The temporal scope of the enquiry extends to the entire lifetime of the settlement, as the words “at any time” make clear. Furthermore, the words in brackets after “at any time”, namely “(whether during any such period as is referred to in section 86(1) below or later)”, confirm, significantly, that if the EBT is one for a limited period, it is also necessary to examine the position after the end of that period. Parliament could hardly have made it clearer that any possibility of using any part of the settled property to benefit any member of the prohibited classes at any time during the lifetime of the settlement (and not just while it remains an EBT) will disqualify the transfer of value from exemption.

Against this background, I now turn to the prohibited categories of beneficiary which it may be convenient to set out again:

Some preliminary points may be made about the way in which this list is structured. First, each of the four paragraphs describes a particular category of beneficiary, using the words “a person who” or a similar description to denote those who fall within it. Secondly, the members of classes (a), (b) and (c) are mutually exclusive, as the words “any other person” in (b) and (c) demonstrate. The draftsman clearly envisages that the paragraphs will be applied in that order, and that the reader will only move from one paragraph to the next if it is necessary to do so. Thirdly, paragraphs (a), (b) and (d) all use the present tense “is”, whereas (c) uses the past tense (“has been”) and embraces people who have been (but by inference no longer are) participators in the relevant company within (a) or (b), either during the period of ten years before the transfer of value or at any time thereafter. It will be noted that paragraph (c) does not include a person who was a participator at the time of the transfer of value itself. The natural inference is that it was unnecessary to do so, because such a person already falls within either paragraph (a) or (b) as a person who “is a participator” at that date.

In examining the position as at the date of the transfer of value, there should be no difficulty in ascertaining whether the actual or potential beneficiaries of application of capital under the settlement (“capital beneficiaries”) include any person who falls within paragraphs (a), (b) or (d). Paragraph (a) asks whether a person is a participator in the company at that date. Paragraph (b) similarly asks whether a person is a participator in a close company which has (in the past) added property to the same settlement by a disposition within section 13 of IHTA. (Section 13 provides that a transfer into an EBT made by a close company is not a transfer of value if conditions equivalent to those in section 28 are satisfied). Likewise, the question whether a person is “connected” with an existing participator within paragraph (a) or (b), and thus falls within paragraph (d), should in principle be easy to answer as at the time of the transfer of value.

The position under paragraph (c) is a little more complex, because it looks forwards in time as well as backwards, but to the extent that the scope of the paragraph is retrospective, it should again be simple to identify any persons who were (but have since ceased to be) participators in the relevant company within the previous ten years. So too, persons connected with any such former participator, as at the date of the transfer of value, should be readily identifiable.

To the extent that paragraph (c) looks forward, it is necessary to ask whether any person who might subsequently become a former participator in the company is capable of being a capital beneficiary. That is obviously not a question which can be answered by providing a list of such people in advance: only time will tell who falls into the category. Accordingly, the only practical solution, if exemption is to be obtained, is probably to ensure that the trusts are drafted in such a way as to exclude such persons from all possibility of benefit, using language which tracks or incorporates by reference the wording of section 28(4). The existence of this limited problem does not suggest to me, however, that it is wrong to begin by examining the position as at the date of the transfer of value. On the contrary, that must be the starting point for the reasons which I have sought to explain, and if such examination shows that persons who are at that date connected with an existing or former participator in the company may in the future become capital beneficiaries, the exemption will be lost. That is in essence the position in the present case, and it yields a result which accords with the clear anti-avoidance purpose of section 28(4) instead of frustrating it.

Further than this, for present purposes, it is unnecessary to go. It is enough that, in my view, any sensible construction of section 28(4) must require the trusts of the settlement to be examined as at the time of the transfer of value, and membership of any of the prohibited categories of capital beneficiaries as at that date will be fatal, whether or not they would still be members at the time when settled property is applied for their benefit. At the very lowest, the risk of this construction being correct was in my opinion so strong that Mr Baxendale-Walker was clearly under a duty to give what Asplin LJ has called a “specific warning” to Mr Barker that his own more sanguine view might well be successfully challenged by HMRC.

For the avoidance of doubt, I emphasise that I am not holding that the date of the transfer of value is the only date at which membership of the four prohibited classes has to be considered. Indeed, I incline to the view that section 28(4) requires the position to be examined throughout the lifetime of the settlement, with the four paragraphs being applied to circumstances as they may turn out to be from time to time. On this approach, the word “is” in paragraphs (a), (b) and (d) should be construed as meaning something like “is from time to time”, or as Mr Furness QC put it in oral argument, to have an “ambulatory” meaning. I prefer, however, not to express a concluded view on the point, because this is not a tax appeal, and we have not had the benefit of submissions from HMRC.

Finally, although this question was barely touched upon in argument, I think it is worth pausing to consider whether the terms of the EBT Trust Deed, as executed on 6 October 1998, may have succeeded in complying with section 28(4) construed in the way I have outlined, even though it was of course the intention of Mr Baxendale-Walker and his firm that the trust property could safely be applied for the benefit of persons connected with Mr Barker during his lifetime after his death. Clause 1.4 of the Trust Deed defined “the Beneficiaries” in terms wide enough to include Mr Barker and his immediate family, but subject to a proviso that “no Excluded Person shall be a Beneficiary”. Excluded Persons were then listed in Schedule 3 as follows:

It seems reasonably clear that this language was intended, at least in general terms, to reproduce the effect (if not the precise wording) of paragraphs (a), (c) and (d) of section 28(4). If, however, that is the correct construction of Schedule 3, a question which fortunately we do not have to decide, and if section 28(4) has the meaning which I would attribute to it, it would then seem to follow that Mr Barker would indeed have been entitled to exemption from IHT on his transfer of shares, but the settlement would never have been capable of operating in the way which he hoped, and on the strength of which he had paid an enormous fee for the tax avoidance advice given to him by Mr Baxendale-Walker. Indeed, it would seem to follow that the sub-trust established soon after, on 23 March 1999, was itself invalid because it was made for the benefit of Excluded Beneficiaries.

In those circumstances, Mr Baxendale-Walker and his firm would be impaled on the horns of an uncomfortable dilemma. Assuming HMRC’s construction of section 28(4) to be correct, he should have given Mr Barker a clear specific warning to that effect; and he could not save the day by arguing that the trusts of the EBT were in fact drawn in a manner which would secure exemption, because the settlement would then fail to achieve the very objective which had induced Mr Barker to make the transfer, and on the strength of which the scheme had been sold to him. Either way, Mr Baxendale-Walker and his firm were clearly negligent.

I agree that the appeal should be allowed for the reasons given by Asplin and Henderson LJJ. There is no real dispute that a solicitor instructed to advise upon and prepare the trust and other scheme documents necessary to minimise Mr Barker’s liability to CGT and IHT would have been required to consider not only whether the s.28 conditions for exemption would be satisfied but also the degree of risk that the construction of the relevant statutory provisions on which the scheme was premised might (if tested) turn out to be wrong.

The degree of risk (looked at through the eyes of the reasonably competent solicitor) will dictate the nature of the warning that the client should be given, although other relevant factors for the solicitor to consider and bring to the client’s attention in relation to a tax avoidance scheme of this kind will be the likelihood of a challenge to the scheme by the Revenue and the degree of judicial and other scrutiny which the scheme will receive.

This case is perhaps more complicated than others because of the judge’s findings on causation in [175]-[176] of his judgment. Mr Barker would have continued with the scheme unless he had received a specific warning that there was a significant risk that the scheme as drafted would not qualify for relief under s.28.

Although it is not contended that it was negligent for the respondents to have taken the view which they appear to have done about the construction of s.28(4), that is not the point. A careful and competent assessment of the rival arguments on construction (which was never undertaken) would in my view have disclosed that there was at the very least a significant possibility that the words “at any time” meant exactly what they said. The judge’s finding that the respondents were not in breach of duty in failing to give the specific warning I have referred to was based on his view that the interpretation of s.28 which they relied on was likely to be correct. With respect to the judge, I disagree. Like the other members of the court, I prefer the construction of s.28 put forward by Mr Furness QC on behalf of the claimant. At the very least, the arguments were finely balanced and, on the judge’s finding, this was enough to have required the respondents, acting as reasonably competent solicitors, to have given Mr Barker the kind of warning which would have led him to abandon the scheme.