Delphi Derivatives Limited v The Commissioners for HMRC

The assessments under appeal are raised pursuant to paragraph 1, which states:

Paragraph 3 defines the ‘Degrees of culpability’, whereby:

Paragraph 4 provides for the standard amount of penalty imposable as a percentage of the potential lost revenue (‘PLR’): (a) 30% of PLR for careless action; (b) 70% of PLR for deliberate and not concealed action; and (c) 100% for deliberate and concealed action.

Paragraph 5 defines PLR as the additional amount due or payable in respect of tax as a result of correcting the inaccuracy or assessment.

In terms of penalty mitigation, paragraph 9 provides for reductions for disclosure in terms of ‘telling HMRC about it’, ‘giving HMRC reasonable help in quantifying the inaccuracy’ and ‘allowing HMRC access to records for the purpose of ensuring that the inaccuracy’ is fully corrected. The ‘quality’ of disclosure includes ‘timing, nature and extent’: sub-para 9(3).

Paragraph 10 sets the ‘standard percentage’ for a penalty imposable according to the degrees of culpability, and the maximum and minimum penalty range (after allowing for mitigation) according to whether the disclosure is ‘prompted’ or ‘unprompted’.

Paragraph 11 provides HMRC with the discretionary power to reduce a penalty because of ‘special circumstances’.

Paragraph 13 provides for the procedural aspects for raising a valid penalty:

Paragraph 14 provides that ‘HMRC may suspend all or part of a penalty for careless inaccuracy under paragraph 1 by notice in writing to P’ by setting suspensive conditions.

Paragraph 15 provides ‘A person may appeal against a decision of HMRC that:

The Tribunal’s jurisdiction on an appeal, so far as relevant, is set out at para 17:

Paragraph 18 (under Part 4 for Miscellaneous) provides for the extent P is to be assessed for culpability where agency is involved. The relevant sub-paragraphs are:

Employee benefit contributions

This section applies if, in calculating for corporation tax purposes the profits of a company (“the employer”) of a period of account, a deduction would otherwise be allowable for the period in respect of employee benefit contributions made or to be made (but see subsection (4)).

No deduction is allowed for the contributions for the period except so far as –

An amount disallowed under subsection (2) is allowed as a deduction for a subsequent period of account so far as –

This section does not apply to any deduction that is allowable –

Clavis Tax Solutions Limited (‘Clavis’), the creator and promoter of the Scheme;

Herald Employment and Recruitment Services Limited, trading as Herald Resource (‘Herald’), a human resources company based in Jersey;

Mark Langran (‘Langran’), director of Delphi, and beneficiary of the Scheme;

James Kelland (‘Kelland’), director of Delphi, and beneficiary of the Scheme;

Bruce Martin (‘Martin’), director of Delphi, and beneficiary of the Scheme;

John Forbes of Forbes Accountants (‘Forbes’), introducer of the Scheme;

Peter Tucker FCA CTA TEP (‘Tucker’), partner of Dickinsons Chartered Accountants (‘Dickinsons’).

By virtue of Schedule 24 to Finance Act 2003 (and later s1290 of Corporation Tax Act 2009), corporation tax relief was denied in relation to contributions paid by an employer company into an employee benefit trust (‘EBT’).

The employer company could claim corporation tax relief when the EBT contributions were applied to benefit an employee, at which point income tax under PAYE and national insurance contributions (‘NIC’) would become payable in respect of the benefits so applied.

The Scheme purported to enable tax-free remunerations of employees via the use of an offshore EBT, and to procure an immediate deduction for corporation tax for contributions made to the EBT and the fees for the use of the Scheme.

The Scheme relied on the exception provision under sub-section 1290 (4)(a) CTA 2009 to claim the CT deduction.

The main feature of the Scheme was to secure an immediate CT deduction, while claiming that no PAYE and NICs arose on the contributions made by the employer to the EBT which would be allocated to sub-trusts designated for the appellant’s directors.

Herald was a human resources company based in Jersey, and would offer a taxpayer company a service in the form of a review for the purpose of making recommendations as to how key employees, such as the company’s directors, ought to be ‘rewarded and incentivised’.

Herald outsourced that service to a UK limited liability partnership by the name Herald Employment Services LLP (‘HES’), whose members included Clavis’ directors.

The findings of HES’ review were to form the basis of the recommendations made by Herald in a report, which would detail various methods of reward such as the payment of a dividend or bonus.

Invariably HES would recommend that rewards be provided by the taxpayer company settling an amount equal to the sum which the review had found would reward and incentivise the directors into an offshore EBT from which the directors could benefit.

Herald Trustees Limited (‘Herald Trustees’) was the trustees of the EBT, and the directors of Herald Trustees were the same as Herald’s directors.

Herald would send an invoice at the same time as the recommendation report for an amount which equated to its fees for implementing the recommendation, plus the sum which it had recommended be made to the EBT; that is, the amount invoiced by Herald was inclusive of the sum to be paid into the EBT.

To implement the recommendation, the company would pay the full amount invoiced directly to Herald. This was supposedly to ensure that the total amount would qualify for the exception under sub-section 1290(4)(a) CTA 2009.

Herald’s fees for each implementation exercise were approximately 10% of the invoice total. The balance of approximately 90% of the invoice total would be settled on the EBT in the company’s name.

Herald would deduct its fees form the invoice total before settlement contribution into the employee benefit trust in the company’s name.

A sub-trust would be set up for each employee who was to benefit from the arrangements, and a share of the balance settled into the EBT would be allocated to each sub-trust.

The funds in the sub-trust would be used to benefit the employee, such as, by making loans from the sub-trust designated for the benefit of the particular employee.

The company would then claim a corporation tax deduction for the payment it had made under the invoice, on the basis that the invoice payment constituted fees paid to Herald under sub-section 1290(4)(a) CTA 2009.

On 17 July 2008, Langran and Kelland attended a presentation by Forbes in respect of the Scheme. Langran understood the Arrangement as a ‘a form of remuneration tax planning’, and would involve ‘advice from a remuneration consultant and the use of an Employee Benefit Trust’.

Also on 17 July 2008, Langran asked Tucker to contact Forbes in relation to the Arrangement. Forbes asked Tucker to sign a non-disclosure agreement (for Probiz), and also confidentiality agreements (from Clavis) on behalf of Dickinsons on 21 July 2008, prior to arranging a meeting for a full presentation of the Arrangement to Tucker.

On 31 July 2008, a meeting was held at Dickinsons’ office in Rickmansworth, attended by Tucker, Forbes, and David Cowen of Clavis, wherein the Arrangement was explained (‘the first/July meeting’).

Following the first meeting, Tucker had a half-hour telephone conversation with Langran and Kelland wherein he explained his understanding of the Arrangement. Dickinsons was then instructed by Delphi to ‘formally review the Clavis remuneration strategy’.

On the morning of 4 August 2008 (‘the second/August meeting’), Tucker met with representatives of Clavis in Jersey, accompanied by David Cowen. Tucker was given sight of the tax opinions of Andrew Thornhill KC on the Arrangement.

On the afternoon of 4 August 2008, Tucker was emailed further documents including: (i) a spreadsheet showing the comparison of various methods of profit extraction from the company, (ii) confirmation regarding the level of Professional Indemnity cover for Herald Trustees, and (iii) a copy of the latest trust instruments.

Several follow-up meetings between Tucker and Clavis’ representatives took place between 4 and 7 August 2008, including one at the trustees’ office in Jersey. Tucker was provided with copies of the draft documentation used for the Scheme.

‘I am instructed by Clavis Solutions Ltd/ Clavis Tax Solutions LLP (“Clavis”) in relation to the use of a Human Resources Consultancy Company (‘HRC’) and a special purpose trust to provide discretionary benefits to employees of UK companies in a similar way to an employee benefit trust.’

Intentions of the Scheme:

The contractual arrangement between the trading company and the HRC:

The role of a HRC is set out by Thornhill in the following terms:

The timing of payment of an ‘overall fee’ is described as follows:

‘The relevant legislation governing the corporation tax deductibility of payments for employee benefit contributions is Schedule 24 to the Finance Act 2003 as amended by draft legislation introduced in the budge on 21^st March 2007. The broad aim of these measure is to restrict deduction for EBT contributions until such time as those contributions are applied to benefit employees in a form which gives rise to an employment income tax charge and an NIC charge.’

‘The Finance Act 2007 will introduce further measures to restrict the availability of corporation tax deductions for contributions to an EBT.

However, the anti-avoidance measures as they currently stand do not apply “in respect of anything given as consideration for goods or services provided in the course of a trade or profession”. See FA 2003 Schedule 24, para 8(a). Clearly, the services envisaged must concern the provision of employee benefits. Otherwise the legislation would not apply. The services envisaged do of course, directly concern the provision of benefits. It is difficult to see how any limitations could be implied so as to bring the services here outside the protection of paragraph 8(a).’

‘In my opinion, therefore, properly implemented, the proposal outlined above should allow the employer company to claim a corporation tax deduction for the amount paid to the HRC. Subject to the points outlined below, I confirm that in my view the payments should fall into the exclusion from Schedule 24 FA 2003 set out at paragraph 8(a) of that Schedule – “the goods and services exemption”.’

‘For the proposal to work, it is vital that the HRC does not act as agent for the company. Payment must be made unconditionally to the HRC for the functions it is to perform and HRC must have complete freedom to then perform those services as it sees fit within the confines of the report. These decisions must be taken – and be seen to be taken – by HRC. In my view, documents confirm that this is indeed what will happen.

The fee paid to HRC will encompass both work to be undertaken by HRC in assessing the rewards to be provided and the cost of making those rewards. Should the situation arise whereby HRC is paid significantly more than HRC ultimately pays out to employees, HRC must be free to retain the excess.’

‘I further confirm that it is my view that the anti-avoidance legislation set out at paragraph 33 of the 2007 Finance Bill further restricting deductions for employee benefit contributions would not affect the proposal set out above.

A deduction would be claimed on the grounds that the payment was made wholly and exclusively for the purpose of the trade. If the payment is made after the year end, it is vital that a liability is established in terms of FRS 12 during the period of account. It does not seem that payment needs to be made within nine months of the year end (see FA 1989 s. 43(2)) though it might be sensible to make it within that time.’

One dated 28 June 2007 advised on ‘the tax effectiveness of providing benefits through the use of trusts and sub-trusts’ in the Clavis Arrangement.

One dated 9 July 2007 was on ‘the use of a Human Resources Consultancy Company to provide discretionary benefits to employees of UK companies in a similar way to an [EBT]’.

A conference meeting between Thornhill and Clavis on 4 October 2007 at Pump Court Tax Chambers in London, was followed by Thornhill’s opinions stating that:

On 3 March 2008, a further opinion was provided by Thornhill on the issue ‘beneficial loans’ with reference to section 720 ITA 2007.

On 10 March 2008, a 5-page opinion was given on ‘further income tax issues relating to the outsourcing strategy’.

The letter starts with: ‘In accordance with your instructions’, and relates the two meetings Tucker had with Clavis Solutions Ltd (Frank Harris and David Cowen on 31 July and 4 August 2008) who explained the structure of the Arrangement in the first meeting, along with ‘a selection of Counsel Opinions’. In the second meeting, Tucker was shown the ‘Instructions to Counsel’, ‘a “Bible” showing the documents which make up the arrangements’, and ‘a file explaining the status and regulation of the Jersey based Human Resources Company’.

This paragraph relates the documents made available to Tucker, noting that he had ‘only been shown the paperwork for a sub-trust’ (and not the main trust document).

The third paragraph is short and contains the caveat: ‘I have not as yet taken up any external references on the scheme with other accountants.’

Paragraph 4 summarises the Scheme in the following terms:

Paragraph 5 goes on to refer to another opinion on DOTAS reporting:

Paragraph 6 sets out the attending features of the Scheme using sub-trusts and the loan mechanism:

Paragraph 7 is a cost-benefit analysis, pitching the costs of entering the Scheme against the anticipated tax savings:

Paragraph 8 summarises the economics of the cost-benefit analysis as setting a minimum entry level:

Paragraph 9 identifies ‘possible areas of risk’ as follows:

Paragraph 10 recommends obtaining a second opinion:

Paragraph 11 contains the caveat that the Scheme is open to challenge from HMRC:

Paragraph 12 relates the promoters’ agreement to fund the first stage of litigation:

Paragraph 13 assesses the chance of success in litigation and reiterates recommendation for advice from independent tax counsel:

Paragraph 14 is the final paragraph with conclusions and a disclaimer:

John Forbes would receive commissions for introducing Delphi.

Tucker remarked on ‘a certain artificiality’ from ‘the very narrow gaps when something was forced through in about 10 days in December on the bridle’, and Cowen reassured Tucker that ‘in terms of your case I would obviously by time-phasing the way we would suggest there will be a much larger gap’.

Tucker then commented on the formulaic nature of the reports:

Tucker’s view on his involvement (as regards reports) if Delphi took the Scheme:

When Cowen tried to impress Tucker on the importance of Sempra as supporting the Clavis Scheme ‘more than anything else’, Tucker arrested the flow by saying:

PT: … I wouldn’t recommend it for any amount less [than £250,000] and I’m actually pushing it to a higher level although there is nothing sort of in it for us because we are not involved in your commission arrangements on this one at all.

PT: … my discussion here I think, there was a guy at this client that originally talked of a figure of £1 million… I’m looking at if they’re going to do it then I’d go in for £2 million, or two and a half.

PT: I don’t know if it is appropriate and I don’t sort of float this as a thought to you to sort of consider but we will be charging our client because we will be getting involved in making sure all the documentation is done at the time and not all pre-signed and the like.

PT: But they’ve got to take a decision on it. … I’ve given them a downside, I’ve sort of gone through, I’ve looked at the retrospective legislation point and said my view on it, it’s a personal view, is that its [sic] below 50% because it’s a specific exemption –

PT: … the chances are less than 50% and much slimmer than if it was some other more complex scheme you had then you could be at risk because –

PT: Back in 03/04, when they blocked employment schemes, they said they would block them all from that date back then and retrospectively block everything. But it’s not retrospective because they had announced in advance but here we’re talking about a corporation tax deduction point.

PT: And not an employment point. … So it’s not – not caught within that.

What happens at the point of employment ceasing:

Cowen then suggested having a ‘fallow year’ between cessation of employment and exit from a sub-trust:

Tucker then queried whether a ‘distribution’ (instead of advancing a loan) could be made out of the sub-trusts on exit:

Tucker ended the phone call with a summary of his understanding:

On 21 August 2008, Delphi’s board of directors held a meeting at which they agreed to set up an employment committee with responsibility for the appellant’s strategy for the establishment of an incentive, reward and retention arrangement for the benefit of its employees. On the same day, the appellant sent a letter to Herald asking for details of the services they provided to which Herald responded on 28 August 2008.

On 1 September 2008, the appellant entered into an outsourcing arrangement with Herald under which Herald agreed to provide certain services specified in Schedule 1 to that agreement, which included: (i) the evaluation of the duties of the employees specified by the appellant; (ii) conducting interviews with the employees and the appellant; (iii) production of a report to the appellant recommending the types of benefits to be provided and their approximate costs; (iv) a proposal for an overall fee which should cover the benefits to be provided to the employees as well as Herald’s costs, and (v) the implementation of the agreed proposals.

On 2 September 2008, David Cowen and Sally Fuller, as representatives of HES (both of whom were also employees of Clavis) attended the appellant’s premises and met with Delphi’s directors.

Herald prepared a report dated 30 September 2008 including an evaluation of all three directors, and recommended that all three directors be provided benefits by either cash bonuses or a special purpose trust and that a budget of £1.8m be made available to reward, incentivise and retain them.

Herald issue an invoice for £1.8m to Delphi on 30 September 2008, the same date as the report.

At a board meeting on 6 October 2008, the contents of the report were considered and Delphi’s directors agreed to settle the invoice.

Delphi paid £1.8m to Herald, of which £1,651,376.15 (after deduction of fees payable to Herald) was settled on the Delphi Derivates Limited Special Purpose Trust (‘Delphi’s EBT’) on 7 October 2008.

The beneficiaries of the Delphi EBT included the three directors of Delphi.

A proportion of the sum £1,651,376.15 was allocated to a sub-fund for the benefit of each of the directors in accordance with the recommendations in Herald’s report.

Frank Harris on behalf of HES (Harris was also a director of Clavis) met with the appellant’s directors on 11 November 2008.

Herald prepared a report dated 19 November 2008 which recommended that the three directors be provided benefits by either cash bonuses or a special purpose trust and that a budge of £3.9m be made available to reward, incentivise and retain them.

Herald issued an invoice for that amount to the appellant on the same date.

The contents of the report were considered at a board meeting on the same date and the directors agreed to settle the invoice.

On 20 November 2008, Delphi paid £3.9m to Herald.

On 24 November 2008, the sum of £3,611,111 was settled on Delphi’s EBT.

On 25 November 2008, allocations to the directors’ sub-trusts were made in accordance with the recommendations in Herald’s report.

Frank Harris on behalf of HES met with the directors on 12 February 2009.

Herald prepared a report dated 20 February 2009. This report was identical to the 19 November 2008 report save that the employee evaluations did not include the profit for the period and the recommended was changed to £1.35m.

Herald issued an invoice for £1.35m to Delphi on the same date.

The contents of the report were considered at a board meeting on the same date and the directors agreed to settle the invoice.

On 24 February 2009, Delphi paid £1.35m to Herald.

On 24 February 2009, the sum of £1.25m was settled on Delphi’s EBT.

On 25 February 2008, allocations to the directors’ sub-trusts were made in accordance with the recommendations in Herald’s report.

On 17 September 2009, Officer Walker received a telephone call from Sally Fuller of Clavis Solutions (Clavis Solutions being the ‘tax advisers to Herald Resource’) in response to Walker’s letter of 15 September 2009. The note of the call (by Walker) recorded the following:

On 21 October 2009, Dickinsons responded to HMRC’s letter of 15 September 2009 under the heading ‘Check of CTSA Tax Return for Delphi Derivatives Limited/ Period ended 30^th June 2008’. The letter opened by referring to Officer Walker’s ‘recent discussion’ with Sally Fuller of Clavis Solutions, and enclosed ‘a complete bible of documents’ said to support ‘the key employee reward and incentivisation arrangement undertaken’ by Delphi in the accounting period ended 30 June 2008, together with the correspondence and minutes contained in the ‘bible’. Dickinsons’ letter also refers specifically to being aware of HMRC’s enquiries into other scheme users:

On 22 October 2009, Sally Fuller emailed Kerry Hall (and two others) at Clavis Solutions under the subject heading of ‘Delphi Derivates new sign up pack for SPT4’:

On 27 October 2009, David Cowen for HES met with the appellant’s directors.

The employee evaluation documents supposedly completed on 27 October 2009 stated the recommended budget to be £3m.

On 31 October 2009, Langran’s short email to Cowen stated in full as follows:

Herald’s report produced on 19 November 2009 recommended a budget of £5.4m.

A letter dated 16 November 2009 by Herald Employment Services LLP (HES in Cheshire) to Christina Kiely of Herald Resource (in Jersey) regarding Delphi, states:

By email dated 20 November from Langran to Pauline Egan of Herald Trustees:

By email dated 23 November 2009 from Langran to Cowen:

On 23 November 2009, Langran emailed Cowen that the invoice would need to be paid in two instalments with the first instalment being loaned back to the directors.

On 23 November 2009, the appellant paid £2.7m to Herald.

On 24 November 2009, £2,522,429 was settled into Delphi’s EBT.

On 25 November 2009, the settled amount was allocated into the sub-trusts.

On 3 December 2009, the appellant paid the second instalment of £2.7m to Herald.

On 4 December 2009, the sum of £2,523,364 (net of fees deducted by Herald) was settled into Delphi’s EBT.

On 7 December 2009, amounts were allocated to the sub-trusts.

‘I am pleased to say that our Counsel [i.e. Thornhill] who in fact appeared for the Sempra Company at the hearing has confirmed to us that their strategy was completely distinct from ours. There were six issues at stake and the one point that HMRC were successful on concerned deductibility within the company. I can confirm that there are no negative implications for the Clavis Strategy. In fact, it is exactly the opposite.’

‘In the Sempra case the company argued that Schedule 24 could not apply as its employees were not beneficiaries of the trust (completely different from ours), rather members of their immediate family were. This argument was rejected by the Special Commissioners and so the contribution was not allowable.

The Clavis SPT strategy works because the invoice raised by the HR Co for the outsourced evaluation process is deductible because of the exemption contained within paragraph 8a Schedule 24 FA 2003, it passes the wholly and exclusively test on general principles and the allocation of funds to sub trusts does not constitute earnings.

‘I would reiterate that the decision was very good news for us and Counsel has confirmed this in conference – we await the outcome of the High Court Appeal which is being heard next month and our Counsel is again leading for Sempra.’

On 15 September 2009, HMRC issued a letter to Delphi to open an enquiry into the 2008 CT return under para 24(1) Schedule 18 to the Finance Act 1988.

On 12 January 2011, HMRC opened an enquiry into the 2009 CT return.

‘I am writing to you because your company has been identified as a current or past user of tax avoidance arrangements promoted by Clavis Tax Solutions LLP and administered by Herald Employment and Recruitment Services Limited. … (The Promoter has claimed that there was no obligation to disclose these arrangements to HMRC’s Anti-Avoidance Group; this claim may be revisited once HMRC enquiries have been resolved.)

‘You and other users of the arrangements should know that HMRC is exploring through the Promoter potential challenges to the arrangements with regard to, amongst other things, the validity of the corporation tax deduction claimed for payments for services and whether or not PAYE and NIC should have been applied to payments received by beneficiary employees of the Company usually by way of interest free loans from an Employment Benefit Trust. HMRC is also looking at similar schemes marketed by other Promoters.’

On 3 December 2012, PAYE Determinations and NICs Decisions were issued for 2008-09. These were appealed by Dickinsons on behalf of Delphi on 6 December 2012.

In November 2013, HMRC issued PAYE Determinations and NICs Decisions in relation to the year 2009-10, which were appealed on 19 November 2013.

‘Further to our telephone conversation you asked that I provide details of payments into trusts/subtrusts to enable you to work out a possible settlement figure for the company.’

‘I confirmed that I had studied the schedule provided and was able to reconcile the amounts, which would enable me to produce indicative calculations of the liability arising. However, I still did not have details of any loans taken by the directors subsequent to 31/3/2009, and I would need this information for the purposes of asking colleagues to calculate any IHT liabilities. Agent will try to get this information, however, in the interim, he will accept computations which exclude any IHT. … Agent also confirmed that none of these loans have been shown on P11Ds and, therefore, there would be no credit available for tax paid for class 1A NIC.’

‘I advised the agent that the questions asked in our letter dated 27/2/15 would need to be answered because, if the EBTSO was not taken up, we need to compile as much information as possible prior to potential litigation of these types of case. Agent understood this and said that, as Clavis Solutions (Mr Cowen) had been appointed to act for the company in respect of the EBT only, they should be providing answers to these questions.’

30 June 2016 – Dickinsons sent Officer Taylor a follow-up email which went unanswered.

22 July 2016 – Directors of Delphi were issued with Code of Practice 9 (‘COP 9’) enquiry letters. COP 9 procedure was for ‘investigations where [HMRC] suspect tax fraud’. Tucker consulted with a number of professional advisers on the response which would best protect Delphi.

26 August 2016 – Dickinsons confirmed to HMRC that Delphi’s directors would not be accepting the COP 9 invitation but would cooperate with their enquiries.

14 September 2016 – Tucker emailed Officer Atkinson waiver letters for Delphi directors to communications to be effected with HMRC by email.

26 September 2016 – Tucker wrote to Atkinson making a settlement proposal on behalf of Delphi, with a schedule of payments, and received an acknowledgement on 27 September that: ‘Following the response received by your client’s concerning my letter dated 22 July 2016, I am not (at present time) in a position to discuss the matter further.’

27 September 2016 – Tucker asked Atkinson to provide a reference number for Delphi to make the first payment on account in respect of the proposed settlement. The reference was provided on 28 September with the caveat: ‘any payment on account is completely without prejudice to the rights of the Board of HMRC to take such action as it thinks appropriate in dealing with this case’.

On 28 October 2016 – Tucker advised that the first payment on account of £1m had been made and receipt confirmed.

On 7 December 2016 – Atkinson emailed Tucker to confirm that tax on investment gain issue was relevant and that he had escalated the matter as settlement needed to be made by 31 March 201z`aa 7.

On 25 January 2017 – a second payment on account of £2m was made to HMRC.

On 9 February 2017 – Atkinson emailed Tucker that HMRC were prepared to proceed on a civil basis, and to request copies of the ‘bibles’ of documents relating to the four tranches of the remuneration planning, and an urgent meeting on 20 or 23 February.

10 February 2017 – the meeting date was confirmed for 20 February 2017 at Delphi’s premises in Colchester.

13 February 2017 – Atkinson confirmed that the basis of the meeting was continuing under COP 9, but HMRC would not be bound by the terms of Contractual Disclosure Facility (‘CDF’).

Company Information sheets and Employee Performance evaluation sheets prepared by Herald Employment and Recruitment Services Ltd (for each transaction i.e. x4);

Documents regarding each payment made to the trust, allocation to the sub-trusts and loan applications/payments made to each director from the sub-trusts again for each transaction;

Also the company bank statements confirming the date each payment was made (x4) to Herald Resources.

The deductions claimed for the tranches of payments to Herald in the CT returns were allowed to stand.

Delphi agreed to pay: (i) the income tax due under PAYE and employer’s NICs on the amounts contributed to Delphi’s EBT under the Scheme, (ii) interest on the PAYE and NICs, (iii) the inheritance tax due of £527,209 on the collapse of the EBT.

The parties agreed that the issue of penalties would not form part of the settlement.

Penalty Explanation Letter;

Notice of Penalty assessment;

Schedule of penalties;

Schedule 1 for careless penalty;

Schedule 2 for deliberate penalty; and

CC/FS10 – Fact sheet on suspending penalties for careless inaccuracies.

Tranche 3 was moved to careless from deliberate;

Tranche 4 has changed from 35% to 38.5% due to HMRC having to use information obtained through third parties to establish the culpable behaviour.

Behaviour which led to the inaccuracy categorised as ‘careless’: ‘Customer’s failure to follow the advice provided by professional advisor to the customer.’

Disclosure was prompted ‘because you did not tell us about the inaccuracy before you had reason to believe we had discovered it, or were about to discover it’.

The penalty range for careless inaccuracy with prompted disclosure is from 15% to 30%.

Reduction for quality of disclosure: maximum for Telling 30%, for Helping 40% and for giving access to records 30%, total 100%.

Penalty percentage after maximum reduction given is at the minimum of 15%.

Special reduction: HMRC do not consider that there are any special circumstances.

Suspension: no amount of penalty can be suspended.

Behaviour which led to the inaccuracy categorised as ‘deliberate’:

Disclosure was prompted ‘because you did not tell us about the inaccuracy before you had reason to believe we had discovered it, or were about to discover it’.

The penalty range for deliberate inaccuracy with prompted disclosure is from 35% to 70%.

Reduction for quality of disclosure:

Penalty percentage after 90% reduction applied to the penalty range of 35% to 70% (equating to 31.5% reduction) to arrive at the 38.5% penalty percentage.

Special reduction: HMRC do not consider that there are any special circumstances.

‘There is now agreement between the parties that the Scheme did not “work” in producing the expected tax saving and the settlement agreement reached reflects the agreed position that PAYE and NICs are due upon the [tranches of] payments made [to Herald].’

Delphi instructed Tucker to review the Scheme, and Tucker’s letter of 7 August 2008 set out his view of the Scheme and its strengths and weaknesses.

Tucker highlighted several weaknesses of the Scheme, including five specific areas of risk, and explained that ‘aggressive tax planning will always be open to attack from HMRC’.

Tucker went on to state that not only could he not formally recommend the Scheme but that he recommended the appellant, before entering the Scheme, to seek independent counsel’s advice on the Scheme.

HMRC’s published guidance on Schedule 24 penalties CC/FS7a sets out what HMRC consider to be ‘reasonable care’, including if a taxpayer is unsure, he must ask his accountant for advice and then once it is received, to follow that advice.

It would appear that the appellant, by asking its accountant to review the Scheme, was unsure about the use of the Scheme. The appellant’s adviser gave it clear advice which Officer Barraclough paraphrased as:

HMRC rejected the appellant’s explanation (as understood by Barraclough), given at the meeting on 16 February 2018 that Tucker’s advice ‘was merely a professional disclaimer’, and that ‘no weight could be attached to that advice’.

For the Scheme to work in theory, a genuine independent review of the appellant by a human resource company as the supposed service provider was required. For the first three tranches HMRC accepted that Herald Resource did undertake a review.

Whilst HMRC are unaware of the content of the discussions between Herald and Delphi, there was sufficient information held to suggest that some form of review had taken place, even if the report for the third tranche was word for word (subject to the amount) identical to the report of the second tranche.

HMRC therefore concluded that for the first three tranches, the culpability is attached to the failure to follow the advice provided to the appellant by its accountant, who had highlighted the weaknesses of the Scheme, and was ‘careless’ within the meaning of Schedule 24.

The appellant has provided information to suggest that a meeting took place on 27 October 2009 between Delphi and David Cowen (a director of Clavis) in relation to the appellant’s remuneration planning.

The review on 27 October 2009 was the third review by Herald in 13 months. The draft report of the review by Herald Resource included a recommendation of a bonus of £1m gross per director with an invoice amount of £3m. These figures were detailed in the signed reviews of each director on 27 October 2009 by Cowen.

On 31 October 2009, (after Cowen’s review), Langran emailed Cowen stating that Delphi would like to put £5.4m through the Scheme, even though Delphi did not have sufficient cash reserves to make the payment. In relation to Langran’s email to Cowen, Barraclough stated in evidence that ‘the predominant weighting was attached to’ this email as ‘the instruction to the avoidance scheme’ in assessing the penalty for tranche 4.

The instruction to Herald did not take place until 3 November 2009.

Delphi agreed with Clavis that they would make one payment of £2.7m on 23 November 2009. After fees, the balance would be loaned back immediately to Delphi’s directors to allow reintroduction to Delphi to provide the funds to make the second payment of £2.7m on 3 December 2009.

In directing Herald through Clavis the amount by which the directors wished to be remunerated, the appellant was aware that an independent review of the appellant did not take place. It is HMRC’s position that the appellant was simply determining the amount by which the directors wished to be remunerated through the Scheme.

The appellant had already used the Scheme three times and so would have been familiar with how the arrangements should work. For the fourth tranche, the documents show that it was the appellant and not Herald Resource which determined how much was to be paid, but the transactions were still presented as if the appellant had followed the correct steps. The appellant would have known what had happened, and this is why HMRC consider the fourth tranche differently from tranches 1 to 3 when considering the behaviour that led to the loss of tax, and have classified it as ‘deliberate’.

In cross-examination regarding whether a ‘genuine independent review’ took place for tranche 4:

In cross-examination on whether Herald was not controlled by Delphi:

In response to the Tribunal’s question on Barraclough’s understanding after the judgment of Hose of Lords in Dextra [2005] UKHL 47 in assessing tranche 4:

To the Tribunal’s question of Barraclough’s understanding of the Clavis Scheme being different from the ‘standard’ EBT or Dextra:

To the Tribunal’s question as to why the legislative exemption would not apply:

In response to Tribunal’s question on tranche 4:

Reply continued to the Tribunal by contrasting tranche 4 with the other tranches:

The enquiries into the CT returns were concerned with the CT deductions:

The case of Dextra and the enquiry being on CT deductions:

That there had been no PAYE enquiry opened:

That the first mention of PAYE and NICs being potentially in point was in 2010:

On aspects of the letter from Tucker to Delphi of 7 August 2008:

The main objective of the Arrangement was to enable a contribution into an EBT to get a Corporation Tax deduction.

If there was a contribution to an EBT and a loan from the EBT to employees, the state of authorities meant that ‘there would be no PAYE and NIC at that time’.

The outcome in Dextra where the company sought a tax deduction as well, Barraclough accepted, at least in part, that taxpayers could rely on Dextra.

The mischief which the Revenue would be hoping to cure was to work round Dextra, and the Scheme was trying to circumvent the effect of the decision of the House of Lords in Dextra ‘to allow for a CT deduction taken immediately’.

The arrangements in the present case were different to a ‘standard EBT’, but he could not explain why if in an ordinary standard EBT, it was permissible not to operate PAYE and NIC, that in the present case, PAYE and NIC needed to be operated.

Sempra had been decided in 2008 and the view was that with regard to the PAYE element, the Revenue had not succeeded at that point in time in the way it had in Rangers.

It was ‘not clear that the writing was on the wall for Rangers’because the Court of Session judgment was not until November of 2015.

McNally, who became Kelland’s financial adviser from around 1990, and had visited Delphi’s premises ‘fairly regularly to meet [Kelland] since 2001’.

McNally in turn introduced Mr Forbes of Forbes Accountants, a member of the Probiz Network. In a meeting in early July 2008, Forbes ‘described’ the Scheme to [Langran and Kelland], and Langran thought it was ‘inherent’ in Forbes’ explanation that he recommended the Scheme, but Langran was conscious of getting Forbes’ advice checked by ‘somebody I know and trust’.

Langran stated in re-examination that he could not recall the exact words of Forbes’ advice, but that: ‘I am sure he would have said something along the lines of the scheme would offer one a chance to avoid tax in a legal manner’.

Delphi’s directors approached Tucker to carry out due diligence, because Tucker (in Langran’s words) is ‘an extremely cautious man’.

Langran stated in evidence that he thought Delphi would have paid Dickinsons for Tucker’s advice.

Tucker visited the trustee’s office in Jersey and attended at least two meetings with Clavis in July and August 2008. (It was accepted by Officer Barraclough that Clavis did hold themselves out to be specialist tax advisers.)

On 31 July 2008, Tucker gave his understanding of the Clavis Arrangement to Delphi’s directors verbally, and no concerns were expressed during this meeting.

On 4 August 2008, Tucker was given sight of the 6 tax opinions by Andrew Thornhill QC provided to Clavis.

Tucker stated in evidence that: ‘my recollection is that we did go through the detail of my letter at that meeting’; he also confirmed that the advice he gave orally, such as during the meeting of 7 August 2008, would not have been phrased differently from the advice given in the letter.

Langran confirmed in cross-examination that the directors ‘were taken to various parts of the letter, as you would expect’; and Langran understood Tucker’s advice as indicating that ‘[Tucker] believed the scheme worked from a legal and tax perspective’.

When asked about the references to the Scheme being ‘most effective’ and being ‘aggressive tax planning’ at the same time, Tucker replied:

As author of the advice, Tucker admitted that obtaining a second counsel’s opinion would not alter the risks identified in the advice letter at all.

As to the real risk, Tucker was asked in re-examination:

‘My normal and usual advice for any such scheme would be to ask that another Counsel Opinion from a barrister other than the original one is obtained if the Promoters of the arrangements are prepared to permit this’, the advice to seek a second opinion was given with a caveat as underlined.’

In Tucker’s witness statement, he stated that disclosure of the full ‘bible’ of documents to another counsel was unlikely, as evidenced by the Non-Disclosure Agreements (‘NDA’) demanded by both Probiz and Clavis.

When asked during examination in chief about the promoters of the arrangements permitting second counsel’s opinion on the merits, Tucker confirmed:

When asked by the Tribunal the wording used in relation to ‘my normal and usual advice’, Tucker confirmed: ‘That would be the firm’s sort of policy, and that would have been something which I would be asked to include in such a letter’. When asked about the aim of a second counsel’s opinion, Tucker stated:

When asked by the Tribunal further in relation to his advice to seek a second opinion, Tucker replied:

Langran stated in evidence that he understood the second opinion caveat to mean:

‘A: [The bible] set out the split by which the arrangements would work, starting with the initial exchange of letters with Clavis/Herald. There was obviously a requirement at that time because it was close to the company’s year end, for the company to cover its Companies Act requirement of having a minute in place of the company to show that it would make –

Q: Sorry, I need to pause you here. The company’s year end is 30^th June, is it not? You said it is close to the company’s year end so you were there meeting on 4^th August, and I am asking myself, in what sense is it close to the company’s year end?

A: I think I need to refresh my memory on the particular document that has been referred to. The actual bible of documents which set out the basis on which the remuneration report would be prepared required a visit by representatives of Herald resources, the HR company, to carry out an evaluation, that subsequently after that report was prepared, it would be sent to the company as part of it. I might not have these all in the right order. The company had to establish a remuneration committee so that there were two directors of the company who would approve any steps taken within the arrangements. Subsequent to that, the bible showed that there would be a trust set up in … Jersey, and sub-trusts for the individual directors …(Italics added)

‘Mark Langran had phoned me and the discussion related to what the profits of the period were, and the objective of these arrangements, as I understood it, were to enable profits to be extracted and the latest management accounts had profits in excess of that period, and that was the way the discussion went. It was not a recommendation of the amount; it was setting out the level of the profits for the period, which could be sheltered should the directors wish to make a payment in that order of magnitude.’

‘Q: … is it not right that the only reason the figure in Clavis’s report changes from the one you see in the draft is Mr Langran’s email where he communicates a higher profit figure?

A: I would say no, because accounts were also communicated.

Q: Right. So the accounts are sent in on 30^th October, and then one day later, we are told, … “Can we do 5.4 million instead?”’

A: … I was not party to how Clavis operated internally, I would not have thought they would have produced a report or varied a recommendation without sight of some accounts to actually support what they were doing.’

Q: So, do you want to revisit your answer that it is likely that the figure changed because of the second email [of 31 October 2009]?

Q: … but previously you said you thought it was likely because of Mr Langran’s email on the 31^st, saying “Can we do 5.4”. Are you now saying it is because of the accounts forwarded on the 30^th? …

Q: Which, the accounts or the email of 31^st?

‘We would have had a conversation and it would have been based on the fact that the accounts profit was in excess of that amount, and Mr Langran would have asked whether such an amount could go through the arrangements. That would have been the nature of the discussion. I was not setting it, and the accounts of the Company would have disclosed profits certainly of that order of magnitude and considerably more.’

A: ‘… When we first came to discussing this we had, well, we still only had the same amount of money, whatever it was £3 million or £2.7, … as available profits and the word here is “available”, we had a lot more money due to us, in around three months’ time. Then Clavis, I do not know how they came to hear about this, but whatever, said “okay, if you have available profits being including the further forward cash flow, you can do this loan back mechanism”. That is why it changed.’

[…]

A: ‘The available profits were always £5.4 million, as dictated by 30^th June accounts. The amount of money available for us to pay an invoice was significantly lower because of the cash flow issue. We had no idea, … that we could do this loan back arrangement so Clavis, David Cowen, …, he knew that the available profits per se, as per the accounts, having done his review, were £5.4 million but he knew we could not – I guess he knew we could not pay that amount of money. He then came up with the – he told Peter we could then do the loan back thing.’

[…]

A: ‘I think you are trying to tie me up in knots. The suggestion never came from Mr Cowen “we should do £5.4 million”. The amount of money is determined by how much money Delphi Derivates made; Mr Cowen had no idea about that. We had to instruct him.

Q: You instructed him that you got £5.4 million profit?

A: Yes, the accounts showed that and/or we had to tell him, yes.

‘Turnover has increased by £7.8m compared to last year as a result of the uncertainty in the market due to the global financial crisis. The directors were able to use their expertise to expose [sic exploit] the volatility in the market and generate substantial gains from the future prices they had in place. The gross margin achieved was 90% in comparison to 83% last year, generating a gross profit of £13.9m.

‘Administrative expenses’, of which ‘Directors emoluments’ was at £11,222,034, (£4,459,025 in 2008) with the emolument to the highest paid director being £3,740,678.

The ‘Analysis of net funds’ showed an opening net funds of just over £1m and a closing balance of £2.277m, of which £1.6m was cash at bank.

Notes to the Balance Sheet as respects ‘Creditors falling due within one year’ included Accruals of £5,558,671, of which £5.4m would be for contributions made to Delphi’s EBT in November and December 2009 (after year-end).

‘Q: … you mentioned … that the PAYE returns were prepared on the basis of the advice received from counsel/Clavis. So, I take it that when preparing those returns you were relying on their advice that no PAYE and NIC liability arises; is that correct?

A:… I did not at the time consider that PAYE applied to such payments, and our tax department produced the P35s based on the regular salary payments and the like which had gone through over the course of the year.’

‘1.18 HMRC’s position at the time was also not supported in the courts. In 2002’s Dextra Accessories v HMRC, the Special Commissioners found that the employee benefit rust (EBT) scheme under consideration achieved the “outcome promised when they were being marketed”. While HMRC was eventually successful in appealing narrower arguments around corporation tax in the House of Lords the question around whether the loans were income was not considered further. It took until 2017 – subsequent to the announcement of the Loan Charge – for Dextra to be overruled by the Supreme Court, which concluded that it had been wrongly decided.

While loan scheme use was growing in early 2000s, HMRC had not yet received judicial support that loan schemes did not work, and that loans should be taxed as income. Taxpayers did not, therefore, have to accept HMRC’s view. Evidence received by the Review consistently supported the view that such schemes were not seen as being aggressive tax avoidance at the time.’

‘Form 2009 onwards, HMRC also published its view in specialist Spotlight articles, which reach a limited number of agents and tax professionals. Evidence about the readership of relevant Spotlights from the time is not available, but in 2015 (when data is available) the four articles published to that point received an average of just 520 views each. The Review therefore concludes that both individual scheme users, and those using schemes through their employers, would likely have continued to be largely unaware of HMRC’s position at this time.’

‘The legal position at the time also remained unclear, with the courts not accepting HMRC’s view of the tax consequences of loan schemes. The 2008 decision in Sempra Metals v HMRC rejected the government’s arguments that loans made by an employer’s EBT were subject to income tax. The case was not appealed to the higher courts, which may have given scheme users at the time a degree of comfort that the legal position was settled. As had been the case with Dextra, it would take until 2017 for the Supreme Court to conclude that Sempra had been wrongly decided.’

‘Spotlight 5: Using trusts and similar entities to reward employees –

PAYE (Pay As You Earn) and National Insurance contributions (NICs), Corporation Tax and Inheritance Tax

We’re aware that companies have been seeking to reward employees without operating PAYE/NICs by making payments through trusts and other intermediaries that favour the employees or their families. The arrangements usually seek to secure a Corporation Tax deduction, as if the amounts were earnings at the time they are allocated, and also defer PAYE/NICs or avoid them altogether. Our view is that at the time the funds are allocated to the employee or his/her beneficiaries, those funds become earnings on which PAYE and NICs are due and should be accounted for by the employer.

In addition our view is that an Inheritance Tax charge may arise on the participators of a close company. Unless the participators are excluded beneficiaries and have not had funds applied for their benefit, such as the receipt of a loan, a charge to Inheritance Tax arises on participators of close companies at the time the funds are paid to the trustee by the close company. Relief is only available to the extent that a deduction is allowable to the company for the year in which the contribution is made. Later payments of earnings out of the trust that may trigger a deduction to the company would not qualify for relief.

Participators affected by this may need to self-assess a liability to Inheritance Tax. There is further technical advice on Inheritance Tax on Contributions to Employee Benefit Trusts on the HMRC internet site.

We are actively challenging examples of such arrangements and considering legislative options to end further usage of these schemes.’ (Emphasis in bold in original replaced by italics here.)

Following the Supreme Court’s decision in Rangers,it is now common ground that the amounts allocated to the directors’ respective sub-trusts are earnings for PAYE and NIC purposes.

The appellant’s PAYE returns for both years did not include any liability for PAYE or NICs in respect of these amounts into sub-trusts, and were therefore inaccurate.

For the 2008-09 return, HMRC submit that the inaccuracy was ‘careless’, and for the 2009-10 return, the inaccuracy was caused by ‘deliberate’ behaviour.

Notwithstanding the different categories of culpability as concerns the two years, Ms Choudhury’s submissions on ‘careless’ behaviour would appear to be generic to both years, and underpin the basis for imposing the penalties, and that the inaccuracy for the year 2009-10 was an escalation from ‘careless’ to ‘deliberate’ (due to the behaviour as reflected by the manner in which the fourth tranche was implemented).

The behaviour which caused the inaccuracy falls to be judged by the standard of a prudent and reasonable taxpayer in its position: Collis (Footnote: ¹).

HMRC submit that the appellant’s reliance on (a) the ‘prevailing practice’ at the time, (b) the advice received from Tucker, and (c) the advice received from others was ‘insufficient’ for holding that the appellant ‘was not careless’.

In relation to the fourth tranche, HMRC submit that ‘there is insufficient evidence that an independent review took place’ prior to the determination of the amounts to be allocated to the Delphi EBT. In the light of the sequence of events leading to the doubling of contributions from £2.7m to £5.4m, HMRC submit that ‘the appellant’s directors must therefore have known that the report provided a predetermined outcome’, and that ‘its only purpose was to provide a smokescreen for the benefit of HMRC’.

Tucker was responsible for Delphi’s participation in the Scheme from a tax advice and compliance perspective. Dickinsons ‘prepared the Corporation Tax returns, ran the payroll of the Company and completed the end of year forms P35, P11Ds as appropriate’.

It is not sufficient for HMRC to demonstrate that the appellant was ‘careless’ in entering into the Scheme, but this is indeed the carelessness alleged by HMRC.

Even if it could be established that the appellant was ‘careless in deciding to enter intothe Scheme, that has no bearing, no causal connection with, the completion of the P35 returns’.

Far from the Revenue establishing that the appellant was careless in the submission of the P35 returns, the unchallenged evidence is that the appellant ‘took all due care in the completion and submission of the same’, including the engagement of Dickinsons, ‘a perfectly capable and competent firm of Chartered Accountants’ to prepare and submit the returns on its behalf.

As to the ‘deliberate’ penalty in relation to tranche 4, Mr Sherry again emphasises that ‘it is not sufficient for the Revenue to demonstrate that the appellant was careless in entering into the Clavis Arrangements’, nor is it sufficient for the Revenue to show that the appellant entered into tranche 4 of the Clavis Arrangements knowing that the tranche had not been executed as it should have been.

Even if it could be established that the appellant knew that the Scheme had not been implemented in the manner purported, there is ‘no causal connection between that knowledge and the completion of the P35 returns’.

It is the appellant’s case that the P35 returns as submitted was in accordance with the ‘prevailing practice’, and the advice received at the relevant time.

Are the penalty notices valid in terms of meeting the relevant time limit?

Were the P35 returns for the years 2008-09 and 2009-10 inaccurate?

Were the inaccuracies ‘careless’ pursuant to para 3(1)(a) in respect of all tranches?

Was the inaccuracy ‘deliberate’ pursuant to para 3(1)(b) in respect of tranche 4?

‘It is commonplace that, if there is a disputed question of fact admitting of only two possible solutions, X and Y, with party A having the burden of proving X in order to establish his case, if A produces credible evidence in favour of X and B produces none in favour of Y, it is very likely that A will win. B must therefore exert himself if he wishes to avoid defeat. But this does not mean that B ever has the burden of proof.

[…] This does not mean that, by traversing the taxpayer companies’ case, the Revenue have taken on the burden of proving fraud.

Naturally, if they produce no cogent evidence or argument to cast doubt on the taxpayer companies’ case, the taxpayer companies will have a greater prospect of success. But this has nothing to do with the burden of proof, which remains on the taxpayer companies because it is they who, on the law as it has stood for many years, are charged with the task of falsifying the assessment.

The contention that, by traversing the taxpayer companies’ version, the Revenue are implicitly setting out to prove a loss by fraud, overlooks the fact that, in order to make good their case, the Revenue need only produce a situation where the commissioners are left in doubt. In the world of fact there may be only two possibilities: innocence or fraud. In the world of proof there are three: proof of one or other possibility, and verdict of not proven. The latter will suffice, so far as the Revenue are concerned.’ (Sub-paragraphs added)

The burden of proof to displace the discovery assessment rests on the taxpayer throughout. Even if the Revenue has adduced evidence in response, that does not mean that the Revenue, by traversing the taxpayer’s case, has taken on the burden of proof.

In the world of proof, apart from the binary outcome of one or the other possibility, there is the third possibility of ‘not proven’ – that is to say, when the tribunal as the fact-finding body is left in doubt of the fact at issue.

Further, if the Revenue produces cogent evidence or argument to cast doubt on the taxpayer’s case, then the taxpayer will have a greater challenge in discharging the burden required to succeed in their case.

‘P is liable under paragraph 1(1)(a) where a document which contains a careless inaccuracy (within the meaning of paragraph 3) is given to HMRC on P’s behalf.’

‘… P is not liable to a penalty under paragraph 1 … in respect of anything done or omitted by P’s agent where P satisfies HMRC that P took reasonable care to avoid inaccuracy (in relation to paragraph 1) …’ (Emphasis added)

Under para 18(1), HMRC have the burden to prove that there is a prima facie case that the P35 returns in question contain a careless inaccuracy; and

By virtue of para 18(3), P (i.e. the appellant) has the burden to satisfy HMRC (and on appeal the Tribunal) that it took reasonable care to avoid inaccuracy.

To construe the defence under para 18(3) literally, the omission of the definite article means that the reasonable care to be taken to avoid inaccuracy is to be exercised in a generic manner, and is not intended to be specific to the inaccuracy in question that has given rise to a potential penalty assessment.

On a purposive construction, if reasonable care is to be exercised to avoid the inaccuracy, that would have presupposed knowledge on P’s part of the inaccuracy in the first place. If P had the knowledge of the inaccuracy (which gives rise to the penalty in question), P should have taken care to remove the inaccuracy altogether. The formulation of taking ‘reasonable care to avoid the inaccuracy’ as a defence would not have made any sense. To stand as a defence against a penalty, the reasonable care that is taken to avoid inaccuracy cannot presuppose foreknowledge of what the inaccuracy in question is going to be.

The statutory wording for the first condition to be met in raising a discovery assessment under sub-s 29(4) before and after 1 April 2010 (Footnote: ⁵) is as follows:

The repealed sub- 95(1) stated as follows:

In Anderson (deceased)(2009) (Footnote: ⁶), which was an appeal against a discovery assessment, Judge Berner considered the test to be applied in relation to ‘negligent conduct’ under the now superseded version of s 29(4) TMA to be an ‘objective’ test by reference to the ‘reasonable taxpayer’:

In Hanson (2012) (Footnote: ⁷) Judge Cannan considered ‘carelessness’ for Sch 24 purposes with reference to ‘negligent conduct’ and adopted the objective test of the ‘reasonable taxpayer’ in Anderson (deceased), and stated that:

Noting that Anderson (deceased) was concerned with s 29(4) TMA, Judge Cannan in Hanson went on to conclude that there is a subjective element in the test of ‘reasonable care’ apposite to Sch 24 provisions:

In Catherine Grainne Martin (Footnote: ⁸)(2014),Judge Redston compared the error penalty regime under Sch 24 with the predecessor provisions under s 95 of TMA, and similarly concluded that the concept of taking reasonable care in the context of Sch 24 penalty regime incorporates a subjective element, and stated that it is ‘similar to the approach taken on “reasonable excuse”’, and so ‘differs from the strictly objective meaning of negligence’ (at [127]). Judge Redston remarked on the absence of ‘reasonable excuse’ provisions in Sch 24, and hence the subjective element should be accorded in the test of ‘reasonable care’ to counter the absence of ‘reasonable excuse’ provisions in Sch 24.

Collis (2011)concerned an appeal against a ‘careless’ penalty under Sch 24, Judge Berner aptly summarised the test as being objective and referable to the ‘reasonable taxpayer’ but taking into account the subjective attributes of the taxpayer in question:

‘Negligence is the omission to do something which a reasonable man, guided upon those considerations which ordinarily regulate the conduct of human affairs, would do, or doing something which a prudent and reasonable man would not do.’

The test of ‘negligence’ in the Blyth sense is referable to a hypothetical person – the prudent and reasonable man – and is therefore primarily an objective test.

The objective test of negligence at s 95 TMA could be mitigated by the ‘reasonable excuse’ provisions at s 118(2) TMA whereby the subjective element was introduced into the former penalty regime.

In the context of the penalty provisions under Sch 24 FA 2008, however, the objective and subjective elements of the test are combined, and the relevant test has been held by case law on Sch 24 FA 2007 to require consideration of the conduct which could be expected of a prudent and reasonable taxpayer (the objective aspect) in the position of the taxpayer in question (the subjective aspect).

‘… even if it is permissible to look to such materials for guidance as to the intent of Parliament in interpreting legislation, the statements in the materials are not sufficient to conclude that the two terms are simply interchangeable. Although there are indications that the change in terminology was not intended to give materially different results (at any rate as regards penalties), Parliament has chosen to use different words and it is those words which must be interpreted. The starting point must be that the term “careless” as further defined as a “failure to take reasonable care” has to be interpreted according to the usual principles of statutory interpretation.’

The date on which the inaccuracy was corrected (where there was no assessment to the tax concerned) was the date of the settlement agreement between Delphi and HMRC; namely: 29 March 2017.

The period for assessing the penalty began to run from 29 March 2017.

The notices of penalty assessments are dated 23 March 2018, and confirmed by Officer Barraclough’s evidence.

On 23 March 2018, Tucker with Langran were sent by email the relevant letter and schedules (also sent by post) to notify Delphi of the Penalty Determinations.

Tucker’s witness statement confirms that the Penalty Determinations were received by post on 3 April 2018.

Condition 1 is that the document contains an inaccuracy which amounts to, or leads to –

an understatement of a liability to tax, …

‘[50] … if HMRC fail to provide any evidence at all to the effect that a s8 notice was served, they will have failed to demonstrate a crucial fact on which their entitlement to a penalty hinges and the FTT will necessarily set aside the penalties charged for alleged failure to comply with that notice.

[51] Where HMRC have given some evidence that a s8 notice was served, it will then be a matter for the FTT to determine whether that evidence is sufficiently strong to discharge HMRC’s burden of proof. The FTT’s assessment of the evidence should take into account the extent to which the taxpayer is disputing receiving a s8 notice. …’ (Emphasis original)

HMRC had opened an enquiry into 2008-09 CT return on 15 September 2009, which was before the expiry date for the 6-year retention period on 5 April 2015.

HMRC had opened an enquiry into 2009-10 CT return on 12 January 2011, which was before the expiry date for the 6-year retention period on 5 April 2016.

The EBT Settlement Opportunity letter sent to Delphi in December 2014 was to intimate that PAYE and NICs were at issue for users of the Clavis Arrangement, and the letter was followed up by another letter dated 27 February 2015.

For the avoidance of doubt, the penalties have not been imposed on the basis that the appellant and/or its advisers took a different view of the law from HMRC at the relevant time.

Rather, the basis for imposing the penalties is that Delphi failed to follow the advice of its own accountant and/or to obtain further advice from independent counsel. In failing to do either, the appellant acted carelessly.

It is HMRC’s case that ‘a second opinion would have highlighted [certain] points not properly considered in [Tucker’s] letter’. The failure to take a second opinion was the carelessness which led to the inaccuracies in the PAYE returns.

In Dextra,the Special Commissioners had not accepted the Revenue’s argument that the payments to the EBT in that case were earnings, and this argument was not pursued on appeal.

Despite losing the argument in relation to income tax in Dextra, HMRC relied on it again in Sempra where the Special Commissioners dismissed an appeal against assessments to corporation tax in respect of payments to an EBT, but again did not accept HMRC’s argument that payments into the EBTs constituted earnings.

HMRC were unable to pursue this argument further in Sempra because the parties settled before the High Court hearing of HMRC’s appeal.

The Supreme Court in Rangers (2017)held that both Dextra and Sempra had been wrongly decided.

Further, the appellant ignores the fact that in both Dextra and Sempra, it was found that the taxpayer company was not entitled to a deduction for corporation tax for the contribution made to an EBT – even if the taxpayer company did not have to account for PAYE in relation thereto.

The appellant’s witnesses, however, accepted that the Scheme was designed with the aim of bringing about both a corporation tax deduction and monies being paid to employees without deduction of income tax and NICs.

As provided by the legislation at the time, the corporation tax deduction could have been secured by accounting for income tax and NICs on the benefits provided out of the contributions to an EBT.

It is therefore incorrect to say that the Scheme was in accordance with ‘prevailing practice’, or (if different) the decided cases at the time, because the appellant was seeking a CT deduction andavoiding the obligation to account for income tax and NICs under PAYE by means of an arrangement that involved more than simply setting up an EBT.

The Scheme therefore could not have succeeded in its purpose of securing a corporation tax deduction without accounting for income tax and NICs, on the basis of: (i) the wording of the legislation, and (ii) HMRC’s published views.

In the context of the appellant’s contention on ‘prevailing practice’, it is worth noting that both PAYE determinations and NIC decisions (December 2012 for 2008-09, and November 2013 for 2009-10) were issued to the appellant in respect of the Scheme long before Rangers was decided.

‘… it seems to me that a practice may be so described only if it is relatively long-established, readily ascertainable by interested parties, and accepted by HMRC and taxpayers’ advisers alike: compare the decision of the Special Commissioners (Dr A N Brice and Mr John Walters QC) in Rafferty v HMRC [2005] STC(SD) 484 at paragraph 114’. [i.e. “a practice generally prevailing had to be a practice, or agreement, or acceptance over a long period whereby the Revenue agreed or accepted a certain treatment of sums in particular circumstances” per Rafferty.]’

R(oao Cartref & Ors) (Footnote: ¹⁷)wherein claims for judicial review were made against HMRC’s enforcement of the relevant legislation as enacted to tackle ‘Disguised Remuneration’, by notifying the claimants’ liability to a tax charge from 5 April 2019, unless a settlement was agreed in advance of that date. The overarching issue for the judicial review claim was ‘whether an incompatibility declaration on human rights grounds is available in respect of the Loan Charge’ introduced by Finance (No 2) Act 2017, Schedule 11. Mr Sherry highlights that the High Court in Cartref made no comment about the interaction between the decided cases in Sempra and Dextra and HMRC’s contrary views.

It is reported in the Morse Report that the Spotlights reached a ‘limited number of agents and tax professionals’, and that scheme users ‘would likely to have continued to be largely unaware of HMRC’s position at this time’ (i.e. from 2009 onwards).

We reject the notion that there should be a different interpretation of the term ‘prevailing practice’ for present purposes to what has been established in case law in the context of s 29(2) TMA. On the contrary, the fact that a Sch 24 penalty is often imposed in tandem with a discovery assessment supports the interpretation that any reliance on ‘prevailing practice’ as a defence against a Sch 24 penalty is referable to the same standard and proof in line with the defence of ‘the practice [being] generally prevailing at the time’ under s 29(2) TMA against a discovery assessment.

Parliament’s intention that there should be consistency in construing s 29 TMA and Sch 24 FA 2007 is evident by the harmonisation of the statutory wording in s 29(4)TMA to that of Sch 24 FA 2007 as discussed earlier, which resulted in the amended wording from 1 April 2010 to ‘[was brought about carelessly or deliberately by]’ that is the current version of s 29(4) TMA.

We have not heard any convincing legal submissions for the appellant by reference to established case law that the notion of ‘prevailing practice’ for Sch 24 purposes should be differently established from that for s 29(2) TMA purposes. Insofar as the appellant seeks to rely on the notion of ‘prevailing practice’, that notion is to be in harmony with s29(2) TMA.

The appellant’s assertion that Sempra and Dextra had established a settled view of the law has no factual basis. The taxpayer in Hoey similarly argued that the state of the law at the time (based on Sempra and Dextra) meant that it was a fair assumption that HMRC would abide by the court’s decisions, even if the decisions had found against HMRC. The UT found that the taxpayer’s argument fell at the first hurdle regarding the state of law, after noting that:

The appellant’s reliance on Sempra and Dextra as the authorities underpinning the purported ‘prevailing practice’ that PAYE/NICs were not payable on EBT contributions is misguided, as concluded by the UT in Hoey at [170]: both Sempra and Dextra were first instance decisions which did not create a precedent, and while the decisions would have been of ‘persuasive value’, ‘there could not be said to be a settled view of the law’.

Further, Sempra and Dextra (wrongly) held that no CT deduction was available (and no PAYE/NICs payable) on contributions to EBTs. Even taking Sempra and Dextra at their face value, the appellant has failed to establish that these two authorities could be construed as giving rise to a prevailing practice that EBT contributions could avoid PAYE/NICs and at the same timeobtain a CT deduction as claimed by the Scheme: see Cockerill J’s judgment in R(oao Cartref) (Footnote: ¹⁸) which highlighted that HMRC had been consistent in their view that EBTs were ineffective in avoiding PAYE/NICs.

The Morse Report, even if admitted as evidence, does not assist the appellant. The paragraphs relied upon by the appellant, as recorded earlier in the decision, are to be read in the context of the whole report. To the extent that the Morse Report sheds any light on the state of the law on the tax treatment of EBTs, it highlights that there was no settled view of the law. For present purposes, we find the UT’s observation in Hoey at [167] apt:

‘This does not imply approval of artificial tax schemes, or of tax avoidance. If the Loan Charge controversy shows anything, it shows what a bad idea participating in such schemes was in the past and will be in the future.’

HMRC are not contending that the Trust or sub-trusts which formed part of the Scheme were not correctly established and/or were a sham, or that any of the payments in question were not actually made by the appellant.

HMRC’s case is that the appellant did not meet the standard of a reasonable and prudent taxpayer because it did not follow the advice it had been given to obtain advice from independent counsel before entering into the Scheme.

HMRC accept that the appellant did not have any tax expertise of its own. Langran’s witness statement referred to several individuals involved in the Scheme: McNelly, Forbes, Clavis, etc. However, it is clear from Langran’s evidence that the only person who was instructed to give Delphi advice and actually gave advice was Tucker.

Tucker’s advice was primarily set out in his letter dated 7 August 2008. HMRC invite the Tribunal to read the letter for itself and form its own view.

At two separate places, the letter of 7 August 2008 recommended that the appellant obtain advice from independent tax counsel.

HMRC’s case is that the appellant did not follow Tucker’s recommendation as clearly set out in his letter, which puts the appellant in a different position from the taxpayer in Hicks (Footnote: ¹⁹)(wherethe taxpayer did follow the advice it had been given). The present case can also be distinguished from Herefordshire Property Company (Footnote: ²⁰)wherein the taxpayer had used a previous scheme marketed by the same promoter and had no reason to seek further advice.

As a final but important point, HMRC submit that it is reasonably clear that the directors were simply concerned with ensuring that they were not entering into any illegal or criminal activity. They were not concerned with whether they were entering into genuine commercial transactions or whether HMRC considered the required tax savings could be obtained. Both Tucker and Langran referred to how HMRC’s view was not the law. Langran said that he would have phrased the question (i.e. instructions) to Tucker as ‘… please examine the scheme to see if it is a legitimate tax avoidance’. Given the context, HMRC submit that it is unsurprising that the appellant chose to ignore the clear recommendation from Tucker to obtain independent advice.

‘In fact, psychological research has demonstrated that memories are fluid and malleable, being constantly rewritten whenever they are retrieved.’

In cross-examination, and in relation to the meaning of the phrase ‘aggressive tax planning’ in the advice letter,Tucker stated variously:

Tucker was clear that any advice he gave orally would not have been phrased differently from the advice given in the letter. There is a conflict with Langran’s evidence on this point because Langran stated that while Tucker’s advice in the letter was to obtain further advice, he had not said this at the meeting: ‘in the meeting no, on the letter, yes’.

Tucker stated that the risk he had referred to in the last paragraph to be ‘a commercial risk’ (which is understood as intending to counter-balance the list of tax risks itemised in the letter that led to the suggestion that the appellant should obtain independent counsel’s opinion) – but there is no reference to the risk being a commercial risk in the final paragraph, only a reference to the possibility of Delphi’s directors wishing ‘to proceed having taken a commercial view’.

In re-examination, Tucker agreed to the proposition put forward in question by Mr Sherry that the real risk with which independent counsel’s advice would help was that Thornhill’s advice, Clavis’s advice, and Tucker’s understanding of the tax consequences of the Scheme were wrong, (which is understood as a suggested interpretation led by the appellant’s counsel of what the ‘real risk’ possibly could have meant in Tucker’s letter) – and this proposed interpretation was put forward for Tucker’s assent to link the ‘real risk’ as put forward with the advice to seek independent counsel’s opinion.

In relation to the minutes of meeting of 23 June 2008, Langran could not recall having had the meeting, while Tucker’s evidence seemed to be concerned with the Companies Act requirement of having such a minute in place. It is not necessary for us to determine the appeal by making a finding of fact in relation to whether the purported meeting did take place in June 2008 as recorded (which would have to had taken place before Delphi was in fact introduced to the Clavis Arrangement in July 2008). Nor do we seek to make any finding that the minutes were retrospectively put in place in order to meet Companies Act requirement (in view of Tucker’s evidence and the possibility that the meeting did not take place in June 2008 as recorded). The unsatisfactory state of the witness evidence not only cast doubt on the reliability of the witnesses’ recall in this instance, but possibly also on the contemporaneity of the minutes.

‘[21] … Such questions are misguided in at least two ways. First, they erroneously presuppose that there is a clear distinction between recollection and reconstruction, when all remembering of distant events involves reconstructive processes. Second, such questions disregard the fact that such processes are largely unconscious and that the strength, vividness and apparent authenticity of memories is not a reliable measure of their truth.’

‘[19] The process of civil litigation itself subjects the memories of witnesses to powerful biases. The nature of litigation is such that witnesses often have a stake in a particular version of events. This is obvious where the witness is a party or has a tie of loyalty (such as an employment relationship) to a party to the proceedings. Other, more subtle influences include allegiances created by the process of preparing a witness statement and of coming to court to give evidence for one side in the dispute. A desire to assist, or at least not to prejudice, the party who has called the witness or that party’s lawyers, as well as a natural desire to give a good impression in a public forum, can be significant motivating forces.

[20] Considerable interference with memory is also introduced in civil litigation by the procedure of preparing for trial. A witness is asked to make a statement, often (as in the present case) when a long time has already elapsed since the relevant events. The statement is usually drafted for the witness by a lawyer who is inevitably conscious of the significance for the issues in the case of what the witness does nor does not say. The statement is made after the witness's memory has been “refreshed” by reading documents. The documents considered often include statements of case and other argumentative material as well as documents which the witness did not see at the time or which came into existence after the events which he or she is being asked to recall. The statement may go through several iterations before it is finalised. Then, usually months later, the witness will be asked to re-read his or her statement and review documents again before giving evidence in court. The effect of this process is to establish in the mind of the witness the matters recorded in his or her own statement and other written material, whether they be true or false, and to cause the witness's memory of events to be based increasingly on this material and later interpretations of it rather than on the original experience of the events.’

‘[22] In the light of these considerations, the best approach for a judge to adopt in the trial of a commercial case is, in my view, to place little if any reliance at all on witnesses’ recollections of what was said in meetings and conversations, and to base factual findings on inferences drawn from the documentary evidence and known or probable facts. This does not mean that oral testimony serves no useful purpose – though its utility is often disproportionate to its length. But its value lies largely, as I see it, in the opportunity which cross-examination affords to subject the documentary record to critical scrutiny and to gauge the personality, motivations and working practices of a witness, rather than in testimony of what the witness recalls of particular conversations and events. Above all, it is important to avoid the fallacy of supposing that, because a witness has confidence in his or her recollection and is honest, evidence based on that recollection provides any reliable guide to the truth.’

The first instruction given by Langran by email dated 17 July 2008 to Tucker was to call Forbes to discuss the Scheme, but the email itself has not been produced.

Langran, on behalf of Delphi, gave instructions to Tucker to review the Scheme, but there is no written record of the instructions given by Langran, who stated in evidence that he had probably asked Tucker by telephone to review the Scheme.

From the transcript of the phone call between Tucker and Cowen, Tucker described his role as reviewing the way they [i.e. Clavis] had put the Scheme together and Tucker accepted this was accurate in evidence.

Tucker informed Cowen that ‘[his] role in all of this is really to review the way [Clavis/Herald] put the scheme together’, and this was stated by Tucker to interject Cowen who was in the flow to impress Tucker of the legal significance of Sempra.

From Tucker’s own statement of his role in this review, and his nonchalance to the significance of Sempra as lending support to the efficacy of the Clavis Arrangement as a tax avoidance scheme, and Tucker’s evidence that he understood his instruction to be to review the ‘technical aspects of a tax arrangement’ (emphasis added) a reasonable inference is that the instructions to Tucker did not include a review of the efficacy of the Scheme from a tax perspective.

That Tucker’s instructions were centred on the review of the documents and legal instruments to implement the scheme is consistent with the concerns Tucker put forward for Cowen to allay: ‘a certain artificiality’; ‘the very narrow gaps when something was forced through … on the bridle’; ‘these reports are too much of a standard’. Tucker’s review involved him looking through the bible of documents, including ‘a full blown Jersey discretionary trust for all nine yards of it’ which made his heart drop.

That Tucker saw his role in carrying out the review was for the purpose of ensuring that Dickinsons could assist Delphi to get the documentary trail in the right place and at the right time is consistent with what Tucker said to Cowen: ‘if I am getting involved with it … we want to make sure that we’ve not spoilt the ship’.

Tucker admitted to Cowen of the fee earning prospect for his firm whereby Dickinsons would assist Delphi with the documentation if Delphi entered the Scheme: ‘we will be charging our client because we will be getting involved in making sure all the documentation is done at the time and not all pre-signed and the like.’

Tucker’s review would appear to cover the economics of entering the scheme: the costs and benefits analysis whereby he reached the conclusion that ‘the costs will kill it’ and there would be no point going in for just £1m given the overheads, but a minimum of £2m or £2.5m; of insurance cover. It is a reasonable inference that in this cost-benefit analysis, Tucker would have taken into account the tax savings the Scheme was supposed to procure in obtaining a CT deduction and no PAYE/NICs payable.

The economics of the Scheme must have included the tax savings having to more than cover the 10% that Clavis/Herald were to deduct from each invoice total. Tucker referred the below 50% possibility of retrospective legislation stopping the CT deduction as ‘a downside’, and remarked on ‘not an employment point’.

The phone call started with the question which Tucker said he could not ‘bottom out’, and the question was concerned with what was going to happen to the monies allocated to sub-trusts at the point of exit.

The concern was allayed by the proposition of a ‘fallow year’ (allegedly from counsel) for a sub-trust after the cessation of employment. The call that started this question finished with Tucker’s summary of his understanding: ‘leave a fallow year and then the loan can be written off without being benefit’.

A reasonable inference from this question that started and finished the phone call and was the dominant chord running through the conversation is that Tucker’s review was focused with the logistics in implementing the Scheme, and the concern behind this question was how the Scheme would work in reality, so that the trust monies would find their way home, so to say, without incurring a tax charge.

The letter was dated 7 August 2008, and was taken in person by Tucker to attend his meeting with Delphi’s directors on the same day, during which the content of the letter was related and discussed with the directors. The meeting went on for an hour and a half, and Tucker then made the phone call to Cowen to help him ‘bottom out’ the issue as how the trust monies were to make a final exit without a tax charge.

The letter started with the caveat that Tucker had not yet taken up any external references on the scheme with other accountants.

The advice remarked on ‘the merit of simplicity’ of the Scheme by using ‘an exemption within anti-avoidance rules’ by way of sub-contracted services to ‘circumvent’ the rules applicable to ‘the direct use of an EBT’ to obtain a CT deduction, as per Thornhill’s Opinion. That the Scheme was trying to utilise an exemption within the anti-avoidance rules applicable to EBTs was clear from the outset to Tucker.

The dispensation from DOTAS reporting, the off-shore features in the Arrangement to bypass tax on loan interest payable by the sub-trusts, the cost-benefit analysis to set the entry level for costs to be outweighed by the anticipated tax deductions – all point towards the purpose of the Scheme as for tax avoidance.

That the purpose of the Scheme was about avoiding tax was clear to Tucker as the adviser and Delphi’s directors as the recipients of the advice. Tucker accepted in cross-examination that the Scheme was intended to save both corporation tax and income tax, as it would not be ‘sensible’ otherwise. Langran confirmed in cross-examination that he had been aware that the Scheme sought to provide a corporation tax deduction without giving rise to an income tax liability: ‘the purpose of the scheme was to try and avoid both taxes’; and agreed that there would be no point in doing it otherwise. We find that it is in view of the avoidance of both taxes that Tucker considered the Scheme as a form of ‘aggressive tax planning’. We reject the interpretation that ‘anyaggressive tax planning’ at paragraph 11 is a reference to some notional scheme and not the Clavis Arrangement. We find that the very context of paragraph 11 means that the description of ‘aggressive tax planning’ was intended to apply to the Clavis Arrangement.

With the tax avoidance purpose being inherent in entering such a Scheme, the letter then outlined the possible areas of risk at paragraph 9, and correctly identified the likely implications for Inheritance Tax, Income Tax and Corporation Tax.

Crucially, the interaction between Corporation Tax and Income Tax is identified at paragraph 9(4) of the letter, although the risk was in anticipating CT planning failing (rather than ‘the “income tax” side’ failing as it turned out to be).

The tax savings anticipated to be delivered by the Scheme was firmly in the cost-benefit matrix, and the listing of tax risks (not commercial risks) in paragraph 10 follows on logically to advise the appellant to obtain ‘another Counsel Opinion from a barrister other than the original one’ – to address the tax risks (not commercial risks). In the context of Thornhill having been referred to in no uncertain terms at paragraph 4 of the advice letter: ‘the opinion of Andrew Thornhill a well respected QC at Pump Court Tax Chambers’, we find that Tucker was very clear and unequivocal that Thornhill’s opinions should not be taken as the last word as regards the exposure to tax risks on proceeding with the Scheme.

While the advice contained in paragraph 10 is bracketed by ‘My normal and usual advice for any such scheme’ and ‘if the Promoters of the arrangements are prepared to permit this’, that does not detract from the fact that the recommendation that the appellant should seek independent counsel’s opinion is unequivocal due to the prospect of litigation. We find that Tucker assessed that the prospect of litigation was close to certain and is contained in the sentence in paragraph 11: ‘any aggressive tax planningwill always be open to attack from HMRC and their current policy is to litigate everything’ (emphasis added). We find that the adoption of the wording ‘will always’ and ‘litigate everything’ was not intended to be read as hyperbole, but was chosen after deliberation by Tucker and partners in Dicksons who had input into the letter, and that the wording was intended to communicate a close-to-certain prospect of litigation for participating in the Scheme.

We find therefore that Tucker’s advice for seeking independent counsel’s opinion stated in paragraph 10 and reiterated in paragraph 13, was firmly set within the context of the close-to-certain prospect of litigation in paragraph 11. There follows more costs analysis by relating the promoters’ undertaking to fund the first stage of litigation, but any onward litigation would be ‘very expensive’ (paragraph 12). We find that both paragraphs 11 and 12 are substantive in detail, and are not included in the letter merely for window dressing purposes to back up a supposed disclaimer. These paragraphs are included to give substance to the kernel of the advice in paragraph 10, and as reiterated in paragraph 13: ‘I would recommend that the matter be put before independentcounsel’.

We find that the written advice to seek independent counsel’s opinion was given with the close-to-certain prospect of litigation in mind, and this finding is consistent with Tucker’s oral evidence: ‘In terms of asking for another barrister to look at the case, it struck me that with the amounts involved it would give additional defence against any attack that the clients had been negligent in entering into any such arrangements.’

The concluding paragraph where Tucker seemed to be equivocal in his position is to be read in the light of the foregoing paragraphs 10 to 13. On that basis, we find:

‘By taking the role of a tax adviser to Mr Hicks in this respect, Mr Bevis has to be judged by the standard of a reasonably competent tax adviser giving advice to a taxpayer on this matter. The advice which Mr Bevis gave was not advice that could have been given by a tax adviser of reasonable competence.’

First, Tucker was prepared to say that the Scheme ‘seemsmost effective’, but was careful in not staking his opinion on the ultimate effectiveness of the Scheme. Tucker was fully aware that the Scheme relied on the novel feature of the service and goods exemption, and in that respect departed from the standard EBTs. Tucker was circumscribed in his endorsement of the Scheme with the verb ‘seems’ to qualify ‘most effective’. Plainly, why Tucker considered that the Scheme only seems to be most effective was because of the ‘possible areas of risk’ as identified at paragraph 9.

Secondly, Tucker’s assessment of the close-to-certain prospect of a challenge from HMRC was grounded on the factual statement in paragraph 11: ‘Enquiries have been raised into … companies which have utilised these arrangements.’ At the relevant time, Tucker was also a serving member of the Council of ICAEW, and of the Tax Committees of the Tax Faculty of ICAEW. We infer that Tucker would have been aware of the Institute’s representations to HMRC on a range of tax issues, including those on the tax treatment of contributions made to EBTs. Tucker’s professional involvement at the Institute level as a council member and a committee member of the Tax Faculty of ICAEW lent extra weight to his evaluation that ‘any aggressive tax planning will always be open to attack from HMRC’. We accord due weight to Tucker’s advice that HMRC’s ‘current policy’ was to ‘litigate everything’ as coming from a professional not only with years of practice experience as a tax specialist and a STEP qualified practitioner, but with additional expertise and considerable institutional knowledge gleaned from years of involvement at the Council and Faculty levels of the ICAEW.

First, that the appellant had taken reasonable care to avoid inaccuracy and we consider this aspect under the terms of para 18(3) defence.

Secondly, that HMRC have not established the causal link that failure to obtain a second opinion was the carelessness which led to the inaccuracies in the PAYE returns.

It is not accepted that the appellant instructed or received advice from Tucker only. Clavis held themselves out to be experts in the field and provided advice on the Arrangement, shared the advice in the six Counsel Opinions, and gave assurances and ‘comfort’ on multiple occasions that the Arrangement was ‘legitimate’ and ‘worked’.

Through discussion with Tucker, the appellant also relied heavily on the six opinions produced by Andrew Thornhill KC who was ‘the respected QC’ and ‘the representative in Sempra and to [Langran’s] knowledge he had beaten, won a victory against HMRC’.

There is nothing in case law authorities that suggests that in order to take reasonable care, a taxpayer has to specifically instruct a professional in writing or otherwise. Mr Sherry said that Bayliss at [15]makes it clear that unsolicited advice and assurances given by one’s advisers and the promoters of a scheme may be relied upon.

Tucker’s letter of 7 August 2008 should be interpreted by reference to what the author and the recipient knew or understood about tax avoidance arrangements and in particular EBTs at the time. (The Tribunal is asked to refer to ‘the findings of the MorseReportin this respect’.) Mr Sherry also said the letter should not be read with the benefit of hindsight or with knowledge of what transpired in the case law many years later.

Referring to Hicks at [287] (which is an impossible paragraph reference (Footnote: ²⁴)), Mr Sherry submits that ‘entering into a scheme is not in and of itself careless or deliberate’.

The suggestion of a second opinion is Tucker’s August 2008 letter ‘is not as clear cut as HMRC have interpreted it’; that ‘there is ambiguity in the letter in that it refers to “my normal and usual advice would be” and then a recommendation qualified by “considering he amount you may wish to place in these arrangements’ (Mr Sherry’s emphasis). It is submitted that the advice given by Tucker ‘was directed to the question of whether or not the appellant should enter into the Arrangements in the context of a commercially driven decision in light of the amounts potentially involved’.

‘One key point’ emphasised by Mr Sherry is that the appellant ‘did obtain independent advice in the form of [Tucker’s] verbal and written advice’; that Delphi did not just rely on the advice from the Scheme promoters.

Further, that Tucker provided a ‘recommendation’ to obtain a second counsel’s opinion was understood to be simply a ‘suggestion’ which ‘did not tell [Delphi’s directors] that they had to do it’. It is submitted that it does not ‘automatically raise the objective standards by which the appellant has to be judged’.

Mr Sherry submits that Tucker as the author of the letter of advice stated in evidence that, in his view, a second counsel’s opinion ‘would give additional defence against any attack that the clients had been negligent in entering into any such arrangements’. Mr Sherry argues that the ‘recommendation’ was therefore reflective of Tucker’s ‘very cautious mindset and represented a “belt and braces” approach’.

Mr Sherry invites the Tribunal to consider: would a reasonable and prudent taxpayer in these circumstances (and one who understood the recommendation to be a disclaimer) have taken additional Counsel’s advice on the proposal? And submits that the answer is clearly, no.

Advice from Clavis – We accept that the appellant drew ‘comfort’ from Clavis’ advice that the Arrangement was ‘legitimate’. We find that to ascertain the legality of the Scheme was of ‘paramountimportance’ to Delphi, as testified by Langran, who spoke of the directors’ concern not to do anything which would jeopardise their Financial Conduct Authority (‘FCA’) registration because the FCA registration was ‘vital’; hence ‘If there was any hint that [the Scheme] was illegal, criminal, bad, naughty, etc. we would not have touched it with a barge pole’. The emphasis of taking advice from Clavis was to ensure the Scheme was ‘legal’, ‘above board’, ‘legitimate tax avoidance’ – as stated by Langran, but that paramount concern for legality did not equate with taking reasonable care to ensure that Delphi did not under-declare its tax liabilities.

Reliance on Thornhill’s opinions – In relation to Thornhill’s Opinion (or his six opinions) being ‘the cornerstone’ as Langran put it for Delphi’s belief that the Scheme was legally and technically sound, we find that this assurance and comfort to be of the same nature as Clavis’ advice.

Legitimate tax avoidance scheme – We accept that the directors of Delphi, like the taxpayer in Bayliss were assured that the Scheme was a legitimate tax avoidance scheme, on reliance of Clavis’ assurance. This does not assist the appellant, however, since the Scheme being ‘legal’ (in the sense that the transactions involved were not a sham) has no bearing on the correct tax treatment for the transactions in question. It is accepted that the appellant took reasonable care to ensure the legality of the Scheme – but that was no proof that the appellant had taken due care to avoid understating its tax liabilities. The central flaw in the submissions for the appellant is to conflate (if not to confuse) taking due care to ensure the legality of the Scheme with taking due care to avoid inaccuracy in stating Delphi’s tax liabilities.

Advice letter read in the light of the Morse Report – We reject the proposition that the reception of Tucker’s advice letter by its recipient should be interpreted with the Morse Report in mind. This proposition is utterly flawed for the following reasons.

Entering into a scheme is not in itself careless – We accept Mr Sherry’s written submission at paragraph 65 on this point, even though we are unable to trace his citation reference to Hicks at [287]. We accept that Delphi took reasonable care to ascertain the legality of the Clavis Arrangement before entering into the Scheme, but that in itself does not prove that the appellant took reasonable care to avoid inaccuracy in terms as required under para 18(3) Sch 24.

That Tucker’s ‘normal and usual advice would be’ – The emphasis of Tucker’s ambiguity in casting his ‘normal and usual advice would be’ makes no difference to our findings of fact as detailed in addressing Issue 3(b). The important finding of fact as regards Tucker’s advice letter is that there was no outright, unqualified professional endorsement for entering the Scheme from Tucker in the context of his view of HMRC’s policy to challenge ‘any aggressive tax planning’. The lengthy submissions for the appellant to circumscribe Tucker’s advice in paragraph 10 of the letter are to suggest Delphi’s case as: (a) somehow qualifying the normal and usual advice by exception, or (b) being contingent upon Clavis’ permission (Footnote: ²⁶). These submissions might have been relevant for submissions on ‘reasonable excuse’ for failure to follow advice. However, the relevant defence here is not whether the appellant had a reasonable excuse for failing to follow advice, and Sch 24 has no provision for ‘reasonable excuse’.

That Delphi did not just rely on Clavis’ advice – we accept that Delphi took due care in instructing Tucker who was suitably qualified to advise them to the extent that was documented in the advice letter. The critical point is that after receiving Tucker’s advice as discussed for Issue 3(b), Delphi took no action in response to the advice.

That recommendation was a ‘suggestion’ – Tucker was the adviser, and the ultimate responsibility of taking reasonable care to avoid inaccuracy rests with Delphi, not its adviser. To downplay the recommendation as a mere suggestion is again a submission that the appellant had a reasonable excuse not to follow the advice.

That recommendation was reflective of Tucker’s very cautious mindset – we have made relevant findings of fact as to Tucker’s competence as an adviser. We are unable to make contrary finding that Tucker’s advice was somehow over-cautious as Mr Sherry submits. We have regard to the fact that Tucker’s advice to obtain independent counsel opinion in paragraphs 10 and 13 was sandwiched by details in paragraphs 11 and 12, and his overall evaluation of the Scheme was well founded, properly reasoned, and factually substantiated as set out under Issue 3(b).

Would a reasonable and prudent taxpayer take additional counsel’s advice? – The question cast by Mr Sherry, while intended to be rhetorical, is addressed here as a factual question. In this respect, we have regard to the long-standing professional relationship between Tucker and Delphi. We find the directors of Delphi to be sophisticated taxpayers, and had formerly considered other tax avoidance schemes. We find that Tucker’s letter did not give unqualified endorsement of the Clavis Scheme and that it was Tucker’s view that ‘any aggressive tax planning will always be open to attack from HMRC’ – the adverb ‘always’ conveyed certainty of HMRC’s challenge on the Scheme that a prudent and reasonable taxpayer would have taken further action to ascertain the tax savings purported to be delivered by the Scheme would be supported by another counsel’s opinion. Further, the certainty of HMRC’s challenge was grounded in the fact that enquiries had already been opened into scheme users as related by Tucker’s letter. We conclude that a prudent and reasonable taxpayer, with the sophistication as the appellant’s directors, and having due regard for its obligation as a taxpayer to deliver documents that were accurate in stating its overall tax liabilities, would have obtained additional counsel’s advice given the certainty of the prospect of challenge from HMRC.

The submissions, at most, establish that the appellant had taken reasonable care to establish the legality of the Scheme before entering into it, but that does not prove that the appellant had taken reasonable care to avoid inaccuracy in its taxpayers’ returns. The flaw in this aspect of submissions is to conflate or to confound the care taken to establish the legality of the Scheme with the reasonable care required to avoid inaccuracy in general for para 18(3) purposes.

The submissions amount to making a defence that the appellant had a reasonable excuse for failing to follow the advice stated in Tucker’s letter. The flaw in this respect is that the statutory defence for Sch 24 purposes is not whether the taxpayer had a reasonable excuse for the inaccuracy that led to a loss of tax, but whether the taxpayer had taken reasonable care to avoid inaccuracy (in general).

‘Applying the test laid by the case of Bayliss, the real question is whether the Company carelessly filed an incorrect return. The focus is therefore on the alleged error in the return and whether the Company was careless in making that error. In this case, as the Company fully relied on the expertise of Clavis (and their Queen’s Counsel Opinions) as promoter of the Arrangements, on its accountants and its auditors, there can be no suggestion that it failed to take care when implementing and carrying out the first three tranches of Arrangements. In light of assurances that the Arrangements were above board as well as delivering remuneration in a CT tax efficient manner, the fact that the Supreme Court decided nine years later that there was a PAYE/NICs liability attached does not demonstrate carelessness in submitting a P35 on the reasonable assumption that income tax was not deductible. … In any event, if the Company had sought a second Opinion, they may have been advised that a Corporation Tax deduction was not available (although this is highly unlikely in light of the prevailing practice and thinking about EBTs at the time) but this would not have informed the Company on how to fill and submit its P35 return. Had the Company sought a second Opinion at the time, the FTT is invited to hold that, in the light of the authorities [i.e. Sempra and Dextra] (which were the extant authorities at the time), such an Opinion was highly unlikely to have differed as to the PAYE position (there would have been no authority to base a contrary view on).’

The UT did not hold that a reasonable taxpayer would have obtained additional advice; rather, it upheld the FTT’s finding that carelessness was established because the taxpayer company Bella Figura Limited (‘BFL’) obtained no advice.

The UT found at [61] that the FTT erred in ignoring two relevant considerations:

In the appellant’s case, HMRC have failed to show the causal link between what they perceive to be misconduct by the appellant and the fact that PAYE should have been deducted from the payments made to the sub-trusts. In other words, HMRC cannot show that the alleged failure to seek a second counsel’s opinion (in light of reliance on advice by Clavis, Thornhill’s opinions, accountants/auditors that the Arrangement would deliver the tax consequences sought) caused the alleged loss of tax.

As confirmed by Tucker’s evidence, the P35 returns were completed by Dickinsons on the appellant’s instructions in accordance with the views of Clavis and counsel’s opinions. HMRC had initially opened enquiries into the appellant’s company tax returns, which is ‘unsurprising’, since ‘the accounts showed an amount being paid by way of directors’ remuneration and the enquiry enabled HMRC to determine what the payment was for’. In turn, a check was not carried out of the appellant’s PAYE returns because HMRC issued determinations based on the information obtained during the enquiries into the appellant’s CT returns.

The appellant’s submission is ‘wrong’ that a second opinion would not have been different from Thornhill’s because:

HMRC submit that a second opinion would have highlighted the following points, none of which were properly considered in Tucker’s letter:

‘In the absence of subsequent reassurances, completion of a tax return on the assumption that the scheme worked might well have amounted to a negligent behaviour. However, in order for s 95 [TMA] to be engaged HMRC would also have needed to show that there was a causal link between the negligence and the errors in the return.’

Given that HMRC has accepted that the transaction was not a sham this would not be a straightforward point: HMRC would probably need to pursue a line of argument that the errors should have been of sufficient concern to prompt the appellant to seek advice from another tax specialist before completing the return, which should (if the adviser had sufficient expertise) have led to the appellant being advised that the scheme did not work either due to the application of s 16A TCGA or for other reasons. However, HMRC put forward no such argument and it is not obvious to us that such argument would have succeeded.’

The Clavis Scheme was founded on the basis that Herald would undertake an independent review of the appellant. HMRC accept that Herald carried out a review for the first 3 tranches (even when the report produced for tranche 3 was identical to tranche 2, save for the amount).

There is insufficient evidence that an independent review took place for tranche 4:

Tucker initially stated that he did not consider the evaluation had changed as a result of the accounts being sent to Cowen but changed his answer on being re-examined.

HMRC submit that the accounts were for the period 30 June 2009 and there is no suggestion that the figures had changed in the period between Mr Cowen attending the appellant’s premises on 27 October and the date the draft accounts were sent to him.

HMRC submit that the appellant, through Langran, gave a direction to Herald as to the amount it wanted to be put through the Scheme. The directors would have been aware that no review was carried out in respect of tranche 4, and the only review carried out had resulted in a different figure of £1m for each director as the proposed remuneration, as seen in the draft report prepared before Langran’s email of 31 October 2009.

Further, or in the alternative, Herald did not give the appellant’s directors any genuine advice about the level of remuneration to be given to them: it merely recommended a sum equal to the figure Delphi had provided to Herald of £5.4 m.

If the review was independent, the appellant should not have needed to tell Herald the sums in which it wanted the directors to be rewarded. The directors would therefore have been aware that Herald did not carry out a genuine remuneration exercise as to what was required to reward and incentivise them.

HMRC rely on ‘insufficient evidence’ that an independent review had been carried out by Herald to prove their case. ‘An insufficiency of evidence cannot assist in proving HMRC’s serious allegation of deliberate conduct’.

Whilst Mr Cowen’s response to Langran’s email of 31 October 2009 was noted by HMRC as not being provided, HMRC are prepared to invite the Tribunal to make the ‘same assumption’ that Cowen ‘blindly and mindlessly complied with [Langran’s] presumed request’.

A decision to follow the suggestion or wishes of the appellant as to the amounts to be contributed to individuals does not automatically mean that Herald did not carry out an independent review, or was not acting independently of the appellant, or not providing the services agreed under the Outsourcing Agreement: Portview at [35(4)(i)-(vii)].

Even if Herald were encouraged or persuaded by the appellant’s wishes, ‘in the absence of any evidence of collusion or sham, Herald had complete freedom to produce/amend its report as it saw fit’ as per the terms of the parties’ agreement and in accordance with Thornhill’s Opinion.

At the time of tranche 4, the most recent extant decision of the courts on the relevance of the circularity of funding and whether it alerted the analysis of expenditure and receipts was the Court of Appeal’s decision given in October 2008 in Tower Cashback v HMRC. (The decision was reversed on this point – after tranche 4 had been implemented – by the Supreme Court on 11 May 2011.) But the circularity argument does not turn the loans from the trust into earnings if they were otherwise not thought to be so in the light of the extant authorities of Dextra (at first instance) and Sempra.

The argument is ‘fatally flawed’ in light of Barraclough’s acceptance that the appellant did not legally control Herald, Clavis or the Trustees who operated the trusts. A change in the draft report does not mean that the appellant or its directors knew of or could in any way control Herald’s actions or omissions.

These allegations do not explain how the alleged lack of an independent review caused an inaccuracy in the P35 return for 2009-10. It is submitted that even if there was no independent review, this would have only affected the CT position: the PAYE/NIC deductibility would have remained the same at the time the P35 was submitted per the applicable case law.

The appellant knew or was reckless that by so doing, it would prevent the CT exemption (i.e. ‘anything given as consideration for services provided in the course of a trade’) from applying to tranche 4, or that

The appellant knew or was reckless that in doing so, it would alter the PAYE position later.

[82] … for there to be a deliberate inaccuracy on a person’s part, the person must to some extent have acted consciously, with full intention or set purpose or in a considered way’,

‘[86] … depending on the precise circumstances, an inaccuracy may also be held to be deliberate where it is found that the person consciously or intentionally chose not to find out the correct position, in particular, where the circumstances are such that the person knew that he should do so….’

‘[63] In our view, a deliberate inaccuracy occurs when a taxpayer knowingly provides HMRC with a document that contains an error with the intention that HMRC should rely upon it as an accurate document. This is a subjective test. The question is not whether a reasonable taxpayer might have made the same error or even whether this taxpayer failed to take all reasonable steps to ensure that the return was accurate. It is a question of the knowledge and intention of the particular taxpayer at the time.’

The enquiry into the CT return for the accounting period ended 30 June 2008 was opened on 15 September 2009, and was a fact firmly in the background at the time when the appellant embarked on implementing tranche 4.

In fact, the conjunction of events meant that Clavis Solutions (as ‘tax advisers to Herald Resource) was responding to Officer Walker’s request in connection with the enquiry into Delphi by way of Sally Fuller’s telephone call to Walker, while at the same time preparing for tranche 4 to be implemented – Fuller’s call to Walker on 21 October 2009 was followed the next day by her email instruction for the sign up pack for SPT4 to be produced.

Dickisnons as Delphi’s advisers, would appear to have been updated by Fuller, as inferred from the opening paragraphs in their letter of 21 October 2009, and Dickinsons by making direct reference to Officer Jones of SI Liverpool co-ordinating the enquiries into other users of the Clavis Arrangement, was fully aware of the scale of investigation.

Around the time of tranche 4 being discussed, HMRC’s Spotlight 5 would have been in circulation since 5 August 2005, and its archived date was 2 November 2009.

Cowen of Herald/Clavis supposedly had prepared an evaluation report dated 27 October 2009 following his attendance at Delphi’s premises on the same date, which recommended rewards of £1m for each of the directors. However, Sally Fuller’s email of 22 October 2009 (which pre-dated Cowen’s visit of 27 October 2009) was to instruct her colleagues to produce ‘new sign up pack for SPT4’ where she clearly stated to her colleagues: ‘They’re doing £3m and will relate to their year ended 30 June 2009’.

The reasonable inference, from Sally Fuller’s instruction email of 22 October 2009, that the figure of £3m was already determined before Cowen’s visit of 27 October 2009 purportedly to carry out an independent review.

Furthermore, the figure per Sally Fuller’s email of 22 October 2009 would appear to be referable to Delphi’s cash position at the time of tranche 4, being £3m (or £2.7m), according to Langran’s evidence.

When Sally Fuller told her colleague ‘They’re doing £3m’ (before Cowen’s visit), the most probable inference of the identity of ‘they’ would be ‘the directors of Delphi’.

Tucker’s evidence was that the tranche 4 payment was made post-year-end, and included in the final set of accounts by way of an accrual.

The reasonable inference is that the management accounts provided to Cowen by email on 30 October 2009 would not have included the £5.4m.

Between the management accounts on 30 October 2009 and the set of accounts filed on 9 June 2010, an accrual of £5.4m augmented the figure for Directors’ emoluments to £11m, (inclusive of the £5.4m invoice paid to Herald), which represents 86% of the ‘Administrative expense’ total of £12.78m for period ended 30 June 2009.

Tucker’s evidence originally stated that the management accounts would not have changed Herald’s recommendation, then changed to state that Herald would not have changed the figure without supporting documents, such as the management accounts.

Langran’s evidence concurred with Tucker’s amended evidence, in that the change of recommendation in Herald’s report from £3m to £5.4m was due to the set of accounts sent on 30 October 2009.

Langran’s statement in cross-examination was that the email of 31 October 2009 was ‘to instruct’ Cowen of the ‘available profits’ from accounting period ended 30 June 2009 because Cowen had ‘no idea about [the company’s profits]’.

Between the email of 31 October 2009 and Herald Employment (Cheshire) informing Herald Resource (Jersey) on 16 November 2009 of the increased amount of remuneration budget, the reasonable inference is that the loan-back arrangements had been agreed to take place for the recommended sum to change from £3m to £5.4m.

Langran’s email to Cowen of 23 November 2009 was to signal to Cowen to arrange for the first instalment payment of tranche 4 to be loaned back to the directors to meet the second instalment payment of the tranche 4 invoice.

We find that the original recommendation of £3m was by reference to Delphi’s cash position at the time, and the figure of £3m was in correspondence to instruct Herald/Clavis for the purpose of producing the ‘sign up pack for SPT4’ as related in Sally Fuller’s email instruction to her colleagues of 22 October 2009.

We find that the figure of £3m had emanated from the directors of Delphi and was determined before Cowen’s supposed review carried out on 27 October 2009.

Herald revised the recommended figure to £5.4m on being ‘instructed’ by Delphi (via Langran’s email of 31 October 2009) of its ‘available profit’.

We accept Langran’s evidence that Cowen had ‘no idea’ of Delphi’s ‘available profits’ without being so instructed.

Langran, in turn, was advised by Tucker, who would have known from the draft set of accounts that Delphi’s taxable profits stood at around £6.5m without any EBT payment invoice.

The circular loan back arrangements were devised to get round the net funds position at year end 2009, standing at just over £2m (cash plus investments), which was very far short of the £5.4m required to reduce operating profit to £1m.

The original recommendation of £3m referable to Delphi’s cash position would have taken into account of what Langran referred to as cash inflow after June 2009.

Without the loan back arrangements, Delphi would only have the cash to make £2.7m (or £3m) and would have to pay Corporation Tax on circa £3.5m instead of £1m.

We find that Langran’s evidence (a) that ‘[t]he suggestion never came from Mr Cowen “we should do £5.4 million”, because ‘Cowen had no idea about that’, and (b) ‘We had to instruct him’ – to be a truthful representation of what actually happened. To that extent, we find that the final EBT payment £5.4m was attributable to a deliberate action of the appellant.

We also find that the contrivance of the loan-back arrangements of the first instalment of £2.7m for the sole purpose of providing funds for the appellant to pay the second instalment of tranche 4, in order to double the overall CT deduction to £5.4m to be a deliberate action.