Roger Joye v The Commissioners for HMRC [2025] UKFTT 664 (TC)

[2025] UKFTT 00664 (TC)Case No TC09547
FIRST-TIER TRIBUNAL
TAX CHAMBER
Hearing Heard on: 8 December 2023Date Judgment date: 5 June 2025
Appeal reference: TC/2022/02367
TC/2022/11608
INCOME TAX – chargeable event gain – calculation of top slicing relief – overpayment relief claim under Case G Schedule 1AB TMA 1970 – whether Self-Assessment Tax Returns filed in accordance with practice generally prevailing –no – appeals allowed
Before
TRIBUNAL JUDGE GERAINT WILLIAMSJANE SHILLAKER
Between
ROGER JOYEAppellantTHE COMMISSIONERS FOR HM REVENUE AND CUSTOMSRespondentDAVID SUMNERSAppellantTHE COMMISSIONERS FOR HM REVENUE AND CUSTOMSRespondentsMr Tim Good of Absolute Accounting Software for the Appellants for AppellantMr Charles Asuelimen litigator of HM Revenue and Customs’ Solicitor’s Office for RespondentsDECISION

Introduction

[1]The Appellants, Mr Joye and Mr Sumners, appeal against the closure notices issued by the Respondents (“HMRC”) under paragraph 7, Schedule 1A Taxes Management Act 1970 (“TMA”).[2]Both Appellants have appealed against the conclusion and amendment in the closure notice they were issued, which required the payment of additional income tax. The additional tax is considered due on the basis that both Appellants incorrectly calculated the amount of top slicing relief (“TSR”) to which they were entitled.[3]The amendments made in the closure notices bring the following amount of tax into charge: Appellant Year Date issued Amount claimed as overpaid Mr Sumners 2016/17 1 November 2021 £40,758.00 Mr Joye 2017/18 18 June 2021 £55,201.63[4]The appeals concern the application of the phrase “practice generally prevailing” (PGP”) in paragraph 2(8) of Schedule 1AB TMA 1980. Background Mr Joye

Background

[5]On 11 July 2018, Mr Joye filed his self-assessment tax return (“SATR”) for the tax year ended 5 April 2018, which included TSR of £10,819.22 in the calculation of the income tax due.[6]On 5 May 2020, Mr Joye made an overpayment relief claim (“OPR”) in respect of TSR. He provided an amended TSR calculation which showed £66,020.85 TSR to be due.[7]On 25 January 2021, HMRC confirmed that an adjustment would be made to Mr Joye’s statement to give effect to the OPR. However, HMRC stated that they were not in agreement with Mr Joye’s interpretation of the TSR calculation.[8]On 9 February 2021, an adjustment of £55,201.63 to Mr Joye’s statement was made.[9]On 17 May 2021, HMRC opened a compliance check into Mr Joye’s OPR.[10]On 18 June 2021, Mr Joye was issued with a closure notice stating his claim contained an incorrect amount of TSR. As such, the refund claimed of £55,201.63 was not due and Mr Joye’s self-assessment statement had been updated to reflect this.[11]On 12 July 2021, Mr Joye appealed against the closure notice on the basis that the calculation of TSR by HMRC did not meet the conditions for constituting practice generally prevailing from sometime in August 2017.[12]On 9 February 2022, HMRC stated their view remained that the original TSR calculation for the year ended 5 April 2018 was calculated in accordance with the practice generally prevailing at the time. HMRC asked if Mr Joye agreed with HMRC TSR calculation.[13]On 23 February 2022, Mr Joye confirmed he did not agree with the Respondents’ calculation. Mr Joye stated the decision of Sally Judges (as representative for the late R Young) v HMRC [2022] UKFTT 77 (TC) (“Sally Judges”) confirms his calculation was correct. Mr Joye set out why he considered that HMRC’s calculation issued on 18 July 2018 could not be construed as reflecting practice generally prevailing. HMRC understood Mr Joye’s reference to a calculation dated 18 July 2018 to be the automatic calculation of income tax due created as part of Mr Joye’s submission of his SATR on 11 July 2018.[14]On 15 March 2022, HMRC issued their view of the matter confirming their view that Mr Joye’s TSR calculation should be in line with practice generally prevailing at the time the return was filed, which is calculated as they had set out.[15]On 24 March 2022, Mr Joye submitted his appeal to the Tribunal.

Mr Sumners

[16]On 12 January 2018, Mr Sumners filed his SATR for the tax year ended 5 April 2017, which included TSR of £53,383.70 in the calculation of the income tax due.[17]On 9 November 2018, Mr Sumners filed an amendment to the 2016/2017 tax return in relation to gift aid carried back from 2017/18. This amendment was processed on 24 January 2019. No amendments were made to TSR, the amendment was entirely unrelated.[18]On 25 March 2021, Mr Sumners made an OPR of £40,758 in respect of TSR. He provided an amended TSR calculation which he considered reflected the statutory basis for calculation of TSR.[19]On 20 September 2021, HMRC confirmed that an adjustment of £40,758 would be made to Mr Sumners’ statement. However, HMRC stated that they were not in agreement with his interpretation of the TSR calculation. The adjustment to Mr Sumners statement was made on the same date.[20]On 25 October 2021, HMRC opened a compliance check into Mr Sumners’ overpayment relief claim. Mr Sumners was informed that HMRC intended to cancel the credit relating to his OPR which had been made to his self-assessment statement for the 2016/17 tax year.[21]On 1 November 2021, Mr Sumners was issued with a closure notice confirming that the credit made to his self-assessment statement of £40,758 on 20 September 2021 had been cancelled. A repayment supplement of £745.35 was also cancelled, creating a total amendment of £41,502.65.[22]On 16 November 2021, Mr Sumners appealed against the closure notice and requested a statutory review. Mr Sumners made reference to the appeal of Marina Silver v HMRC [2019] UKFTT 263 9TC (“Silver”), which was heard on 18 February 2019, with the decision released on 18 April 2019.[23]On 21 December 2021, HMRC provided clarification about the withdrawal of their appeal to the Upper Tribunal for the decision in Silver and reiterated the reasons for denying Mr Sumners’ overpayment relief claim.[24]On 14 January 2022, Mr Sumners stated his intention to contest the TSR calculation. Mr Sumners provided a computation to demonstrate how the calculation should be performed on the basis adopted in Silver.[25]On 10 February 2022, HMRC issued their view of the matter confirming their stance that Mr Sumners’ TSR calculation should be in line with practice generally prevailing at the time the return was filed, which had not changed for the year in question. The matter was referred for a statutory review. The review conclusion, issued to Mr Sumners on 22 March 2022, upheld the original decision. Mr Sumners submitted his appeal to the Tribunal on 11 April 2022. Issue in dispute[26]The issue in dispute in both appeals is whether HMRC are correct to reject both Appellants’ overpayment relief claims under Sch 1AB TMA (Case G) on the basis that the SATRs (in respect of the TSR claims) were filed in accordance with practice generally prevailing at the time. HMRC accept that s2(8)(a) of Sch 1AB TMA is satisfied and that the Appellants overpaid tax.

Relevant Legislation

[27]The relevant legislation in this appeal is paragraphs 1 and 2 of Schedule 1AB to TMA 1970, s461 onwards of the Income Tax (Trading and Other Income) Act 2005 (“ITTOIA”) and s23 onwards Income Tax Act 2007 (“ITA”). TMA 1970 Schedule 1AB RECOVERY OF OVERPAID TAX ETC 1(1) This paragraph applies where–(a) a person has paid an amount by way of income tax or capital gains tax but the person believes that the tax was not due, or(b) a person has been assessed as liable to pay an amount by way of income tax or capital gains tax, or there has been a determination or direction to that effect, but the person believes that the tax is not due. (2) The person may make a claim to the Commissioners for repayment or discharge of the amount. … 2(1) The Commissioners are not liable to give effect to a claim under this Schedule if or to the extent that the claim falls within a case described in this paragraph (see also paragraphs 3A and 4(5)). … 2(8) Case G is where– (a) the amount paid, or liable to be paid, is excessive by reason of a mistake in calculating the claimant's liability to income tax or capital gains tax (other than a mistake in a PAYE assessment or PAYE calculation), and (b) liability was calculated in accordance with the practice generally prevailing at the time.

ITTOIA

[28]Section 461 provides for the charge to tax. Income tax is charged on gains treated as arising from polices and contracts to which the Chapter applies. Sections 530 to 538 set out the provisions relating to tax treated as paid on gains and to reliefs. Section 535 provides for top slicing relief. It requires a comparison of 1) the amount of the individual’s total liability to income tax on the chargeable event gain (assuming that it is the highest part of the individual’s total income for the tax year) (TL) less the amount of basic rate tax treated as having been paid on the gain (BRL) and 2) the individual’s relieved liability under section 536.[29]Section 536 sets out a series of steps for determining the individual’s top slicing relieved liability for the purposes of s535(1) for tax on a gain from one chargeable event. ITA[30]Section 23 provides the basis of calculation of income tax liability. The statutory provisions require a number of steps to be performed: Step 1 – Identify the amounts of income on which the taxpayer is charged to income tax for the tax year. This represents total income and each amount is a component of total income. Step 2 – Deduct from the components the amount of any relief to which the taxpayer is entitled in the tax year. Total income less deducted relief is net income. Step 3 – Deduct from the amounts of the components after step 2 any allowances to which the taxpayer is entitled for the tax year. Step 4 – Calculate the tax at each applicable rate on the amounts of the components left after step 3. Step 5 – Add together the amounts of tax calculated at step 4. Step 6 – Deduct from the amount of tax calculated at step 5 any tax reductions to which the taxpayer is entitled for the tax year (including top slicing relief). The result is the taxpayer’s liability to income tax for the year.[31]Section 25 at subsection (1) supplements the provisions about reliefs and allowances in Steps 2 and 3 of the calculation in s23. Subsection (2) requires that the reliefs and allowances at steps 2 and 3 of the section 23 calculation be deducted “in the way which will result in the greatest reduction to the taxpayer’s liability to income tax”.[32]Section 35 outlines the rules for entitlement to claim a personal allowance[33]Section 58 provides the meaning of “adjusted net income” (“ANI”) and the steps for calculating an individual’s ANI for a tax year.

Evidence

[34]The documentary evidence comprised a hearing bundle of 548 pages which included copies of correspondence between the parties, the SATRs, HMRC’s SoC and Mr Silver’s speaking note from Marina Silver v HMRC [2019] UKFTT 0263 (TC) (“Silver”) appeal, copies of two letters from Sally Judges, correspondence provided by Mr Good between agents or taxpayers and HMRC, HMRC correspondence with Mr Good , various articles, publications and examples of TSR using HMRC’s test case generator and archived copy of IPTM3850 dated 21 May 2016. There was no witness evidence. Parties submissions Appellants’ submissions

Parties submissions

[35]The appeals concern the application of the phrase “practice generally prevailing” in paragraph 2(8) of Schedule 1 AB TMA. The factual question that the Tribunal is being asked to determine is the date from which HMRC’s incorrect calculation of TSR under sections 530, 535 and 536 ITTOIA ceased to be PGP if indeed such calculations were ever within the definition of PGP.[36]The Appellants contend that HMRC’s incorrect calculation did not fall within the definition of PGP from 6 April 2010 (the introduction of the tapered personal allowance); alternatively from 6 April 2016 ( the introduction of the dividend tax allowance and the personal savings allowance); alternatively from 28 September 2017 (the date of an article “It’s all gone Pete Tong” in Taxation Magazine); alternatively from 11 October 2017 (the date of an email from Darryl Wall of HMRC); alternatively from 6 February 2018 (the date of a letter from Robby Wells of HMRC); alternatively from 29 June 2018 (the date of an invitation to meet with HMRC); alternatively from 12 July 2018 (the date of the meeting with HMRC); and in any event before the relevant dates for the Appellants (11 July 2018 for Mr Joye and 9 November 2018 for Mr Sumners). The Tribunal is requested to make a finding of fact as to the date from which HMRC should not be able to deny overpayment relief claims for wrongly calculated TSR on the grounds of PGP.[37]It is common ground that HMRC’s computerised calculations for TSR were, in some cases, found to be incorrect following (and possibly even before) changes to the income tax legislation from 6 April 2016. In particular the calculation of the available personal allowance and the application of so-called beneficial ordering in the calculation of tax on a single annual equivalent were not correctly coded in the HMRC tax calculator program, as found in Silver. HMRC appealed Silver to the UT and on 13 March 2020 sought consent to withdraw that appeal and agreed to pay the taxpayer’s costs. The Tribunal further found for the taxpayer in Sally Judges in a decision released on 18 February 2022 following a hearing on 19 January 2022.[38]The Silver appeal concerned a tax liability arising in 2015/16 and a disagreement over the tax calculation that began in May 2017. The Sally Judges appeal concerned a tax liability arising in 2017-18 and a disagreement over the tax calculation that began in April 2019. The decisions in Silver and Sally Judges clearly establish that HMRC’s practice in calculating top slicing relief was wrong in law and had been wrong since 6 April 2016. HMRC’s case is that the incorrect view was generally prevailing until the somewhat arbitrary date of the Chancellor’s Budget speech on 11 March 2020, that date cannot logically be the date on which the previously incorrect application of the predecessor legislation ceased to be PGP. An officer of HMRC, Mr Luke Edwards, in correspondence with agents stated that the incorrect practice was generally prevailing until the equally arbitrary date of the decision in Silver, 18 April 2019. It is submitted that the hearing date (18 February 2019) or the date of Mr Silver’s Tribunal submission (8 March 2018) are better but not necessarily the best candidates for the date from which the HMRC methodology ceased to be PGP.[39]In Revenue and Customs Commissioners v Household Estate Agents Ltd [2007] EWHC 1684 (CH) (“Household”), Henderson J (as he then was) stated: “Without attempting to give an exhaustive definition, 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”.[40]In Boyer Allen Investment Services Ltd v HMRC [2012] UKFTT 558 (TC) (“Boyer”), Judge Berner further considered the elements identified by Henderson J in Household: “(1) to be ascertainable required that the practice was not inchoate and that it be sufficiently precise and devoid of uncertainty in its application [34]; (2) although a published statement of practice was the paradigm of an ascertainable practice, it was possible for a practice to be ascertainable if settled, defined and agreed between, or communicated between, taxpayers or otherwise sufficiently identified to the outside world [35]; (3) a published practice was likely to be capable of being regarded as having become generally prevailing over a shorter period than one merely established in practice [36]; (4) an internal practice of HMRC would not be generally prevailing until it could be identified with reasonable clarity and precision by taxpayers [37]; (5) that [the same] quality of clarity and precision must be present in the understanding of HMRC and taxpayers alike [38]; (6) in order for the practice to be "generally" prevailing it must have been adopted by HMRC and generally, but not universally, by the taxpayer community [38]; and (7) the practice would not be settled if it was not applied in a consistent manner.” [40].[41]In John Hargreaves [2019] UKFTT 0244 (TC) (“Hargreaves”), the Tribunal held that the burden of proof fell on the taxpayer to show that his tax return was submitted in accordance with PGP. The Tribunal noted the existence of various articles that raised questions over the published HMRC practice and the Tribunal concluded that the return had not been made in accordance with PGP. The Upper Tribunal in John Hargreaves v HMRC [2022] UKUT 00034 (TCC) (“Hargreaves UT”) noted that Mr Hargreaves sought to discharge the burden by reference to just four documents despite it being clear from the Tribunal decision at [70] to [99] that the Tribunal had engaged in detailed consideration of numerous contemporaneous documents and publications without saying anything about the significant other evidence that weighed heavily in the balance in leading the Tribunal to a different conclusion.[42]It is submitted that the context in which Judge Berner used the word “inchoate” (i.e. in applying Henderson J’s concept of “readily ascertainable”) the most appropriate synonyms are: basic; crude; elementary; rough and ready; and undeveloped. Therefore, until some time after 11 March 2020, HMRC’s computational methodology and hence practice met the definition of “inchoate”.[43]The process through which the actual computational methodology used by HMRC can be ascertained is by analysis of HMRC’s “Test case generator” (“TCG”), the TCG is the underlying formulae that is updated each tax year and embedded in software used to calculate TSR. Officer Keith Graham confirmed at a meeting with Mr Good on 12 July 2018 that the TCG did not conform to the procedure set out in s535 and s536 ITTOIA as the legislation requires a full calculation of tax on the basis that the gain is limited to the annual equivalent, per [30] Silver. Rather than performing the hypothetical calculation required by s536 ITTOIA the TCG adopted a shortcut approach. This shortcut approach is articulated in the archived 5 August 2016 version of IPTM3840, it is submitted that such an approach falls within the definition of “inchoate”.[44]The various published articles concerning HMRC’s calculation methodology show a growing difference between HMRC’s practice and the methodology preferred by commentators and some in the accounting and tax profession. It is submitted that HMRC’s methodology ceased to be PGP from at least 28 September 2017 following an article in Taxation dated 28 September 2017 or, alternatively, 11 October 2017 following e-mails between Mr Cook and HMRC. It is further submitted that HMRC’s methodology ceased to be PGP on 12 July 2018 following a face to face meeting between Mr Cook and HMRC.[45]Numerous postings on AccountingWeb referred to TSR including an article, requests for advice and referencing longstanding issues. The copy correspondence between some 50 odd agents and HMRC is indicative of a wider disagreement amongst agents and taxpayers with the HMRC practice over the relevant period and supports the Appellants’ argument that the HMRC practice was not “adopted … generally … by the taxpayer community” over the relevant period.

HMRC’s submissions

[46]HMRC may refuse an OPR claim if it falls within a case in Paragraph 2 of Schedule 1AB TMA. Case G at paragraph 2(8) applies where excessive tax is paid because of a mistake in calculating the tax liability and where the tax liability was calculated in accordance with the PGP at the time the return was filed. It is not disputed that the condition at paragraph 2(8)(a) is met. HMRC’s case is that the condition at paragraph 2(8)(b) is also met because the TSR calculations (and therefore the tax liability) was calculated in accordance with the PGP on the date each return was filed and the OPR claims fail.[47]In order to establish if the Appellants were eligible for OPR the Tribunal will need to determine if the TSR calculation in each SATR was in accordance with PGP on the date the SATR was filed. The primary dispute between the parties is about the PGP on the date each return was filed. There is no dispute about the validity of the closure notices but the TSR calculations provided by HMRC should be preferred over those submitted by the Appellants.[48]Silver was an appeal to the Tribunal which considered how a taxpayer’s personal allowance should be used within a TSR computation in circumstances where the taxpayer loses their entitlement to the personal allowance within the main income tax calculation at s23 ITA.[49]In Silver, HMRC argued that if s35(2) ITA applied to restrict the Appellant’s entitlement to the personal allowance in the main income tax calculation, the same restricted amount of the personal allowance should be used in the TSR computation; see Silver at [12]. The Tribunal at [30] preferred the Appellant’s interpretation of the legislation and HMRC’s view was inconsistent. The following points about Silver are relevant to these appeals:(1) The Appellant’s tax return and TSR computation was not made with the assistance of an accountant, solicitor, or similar tax professional (confirmed by verbal statement of Mr Silver at the hearing);(2) The Appellant was not represented at their appeal by an accountant, solicitor, or similar professional with tax expertise;(3) HMRC charged the Appellant a penalty for a careless inaccuracy because it concluded that her return was inaccurate and not in accordance with the PGP at that time; and(4) HMRC appealed Silver to the UT but withdrew after new legislation was announced.[50]It is submitted that Silver confirms the PGP up to, at least, the publication of the Tribunal decision on 18 April 2019. The PGP was to restrict the personal allowance per s35(2) ITA and use the restricted personal allowance figure when calculation TSR. If that was not the position:(1) The Tribunal would not at [31] have made findings about Parliament’s intent with TSR; and(2) HMRC would not have charged inaccuracy penalties nor appealed the decision to the UT.[51]It was relevant that the Appellant in Silver did not rely upon a tax professional to prepare the TSR computation and tax return and was not professionally represented, this suggests that the position advanced by the Appellant was not widely held by tax professionals involved in this area of tax law. Published statistics show that 2,000 taxpayers were affected by the s35(2) ITA 2007 restriction of the personal allowance but HMRC are not aware of any appeals, prior to the publication of Silver which challenged HMRC’s view of how the TSR legislation applied demonstrating that the position in Silver, which Mr Sumner argues should be applied to his gains, was not PGP at the time he filed his return. The decision in Silver and the subsequent legislation upended the longstanding PGP when calculating TSR. Both Appellants filed their SATRs before the Silver decision, 18 April 2019, which is the earliest dated that the PGP could be said to have ended.[52]When considering PGP, the relevant date is the date that the return was filed, s29(2) TMA. The question of whether there is a PGP is a question of fact, per Rose Smith & Co. v Commissioners of Inland Revenue (1933) 17 TC 586 (“Rose”) at [593]. For a practice to be described as a PGP it must per Household at [58] be “relatively long-established, readily ascertainable by interested parties, and accepted by HMRC and taxpayers’ advisers alike”. Boyer at [34]-[42] set out the principles for a practice to be considered PGP which HMRC summarised as follows:(1) It must be readily ascertainable with sufficient precision and devoid of uncertainty in the application.(2) It can arise not simply on published statements but also through the development of an understood practice even if unpublished.(3) It must be identifiable with reasonable clarity and precision by taxpayers or their advisers.(4) The clarity and precision must be present in the understanding of both HMRC and taxpayers or their advisers.(5) The practice must have been adopted by HMRC and generally by the taxpayer community.(6) There must be consistency in its application.[53]Neither Appellant has demonstrated that the calculation of TSR in their SATR were not in accordance with the PGP on the date the SATRs were filed, the revised TSR calculations were based on the decision in Silver and legislation introduced on 11 March 2020. None of the requirements in Boyer were met by the revised TSR calculations. The evidence is that there was a settled practice for computing TSR at the time each relevant return was filed, which could properly be described as PGP. This settled practice was not to recalculate the personal allowance in the TSR calculation where a taxpayer had lost entitlement to the personal allowance in the main income tax calculation. The TSR calculations in the Appellants’ original returns, which were prepared by or with the assistance of experienced tax professionals, were accurate and were made in accordance with this settled practice. The Appellants have not explained how their original SATRs included TSR calculations which they say were not PGP at the time the returns were filed nor why they did not seek to correct their returns until after the publication of Silver and the announcement of new legislation. HMRC submit that the most probable explanation for the Appellants’ inaction is that they accepted that their returns were accurate and in accordance with PGP when they were filed and are now seeking to apply the “post-Silver” PGP to “pre-Silver” returns. Paragraph 2(8) of Schedule 1AB to TMA 1970 explicitly prevents such practice.[54]The letter from tax professionals, published articles and meetings with HMRC employees do not demonstrate that the original SATRs were not computed in accordance with the PGP at the time the original SATRs were filed. Where a sufficiently precise practice has been adopted by HMRC and the taxpayer community generally, it can be considered PGP. The Appellants’ suggestion that taxpayers and their agents opposed the prevailing practice but were unaware, unwilling, or uninformed of their right to challenge that practice is unevidenced. The only plausible conclusion is that there were no substantial challenges because taxpayers and their professional advisers and agents understood, adopted and applied what was the PGP even if they felt that the practice was unfair.

Discussion

[55]Both parties’ submissions focused on the meaning of a PGP and we start by considering where the burden of proof lies in establishing the existence of a PGP.

PGP burden of proof

[56]The parties’ position in respect of the burden of proof is as follows: Mr Good for the Appellants submits the burden falls upon HMRC as they are relying upon the PGP and Mr Asuelimen for HMRC submits the burden falls upon the Appellants as they are disputing the quantum of the assessments. The closure notices concern claims by the Appellants for OPR made after submission of the Appellants’ SATRs on the basis that the TSR in the SATRs was not calculated in accordance with the decisions in Silver and Sally Judges. Henderson J in Household, a case relied upon by HMRC, stated the following in respect of where the burden of proof lay in evidencing a PGP:
“[45] … However, it seems clear to me as a matter of general principle that the burden of proof must rest on the party who asserts that there has been an operative mistake in the return, and that the return was in fact made in accordance with the generally prevailing practice. That party will inevitably be the taxpayer, not HMRC. In other words, the burden lies on the taxpayer to establish that paragraph 45 applies, not on HMRC to establish that it does not apply. [46] I base this conclusion on the structure and wording of paragraphs 42 to 45, and on the general principle that the legal or persuasive burden of proof ‘lies upon the party who substantially asserts the affirmative of the issue’: see Phipson on Evidence (16th edn), para. 6-06. The matter can usually be tested by asking which party would succeed if no evidence were adduced on the issue: see, for example, in a tax context, the illuminating judgment of Slade J in IR Commrs v Garvin (1979) 55 TC 24, at 51F–57E, to which I was helpfully referred by Mr Woolf, especially at 54I–55A. In the context of paragraph 45, if no evidence were adduced as to the existence of an operative mistake in the return or as to the existence of the generally prevailing practice, there would be no basis upon which the Commissioners could conclude that paragraph 45 applied, and accordingly nothing to restrict the power of HMRC to make a discovery assessment if the conditions of paragraph 43 or paragraph 44 were satisfied. [47] … the question of the existence of a settled practice will depend on the evidence of taxpayers and their professional advisers at least as much as on the practice of HMRC. … [58] … the position is in my view straightforward. If the company wished to rely on paragraph 45 at the hearing before the Commissioners, the burden was on the company to establish both an operative mistake in the return and the practice generally prevailing in August 2000. The company failed to adduce evidence on either of those questions, and relied only on the submissions recorded in paragraph 6 of the case stated. Those submissions refer to what was alleged to be ‘the professionʼs view’ that section 43 of the Finance Act 1989 did not apply to contributions to EBTs. However, without any evidence to support that assertion, and without any evidence that the Revenue took the same view, there was no material before the Commissioners which could support a conclusion that a settled practice existed, let alone a settled practice which could properly be described as ‘the practice generally prevailing at the time’. Without attempting to give an exhaustive definition, 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 taxpayersʼ advisers alike”: compare the decision of the Special Commissioners (Dr AN Brice and Mr John Walters QC) in Rafferty v R & C Commrs (2005) Sp C 475, at paragraph 114.”
[57]Whilst noting that Henderson J in Household was considering PGP in the context of a discovery assessment pursuant to paragraph 41 of Schedule 18 to the Finance Act 1998, we consider that the principles in respect of the burden of proof of establishing a PGP are equally applicable here in the context of Paragraph 2(8) of Schedule 1AB to TMA. Household is a decision of the Hight Court and is binding on this Tribunal. It is accepted by HMRC that paragraph 2(8)(a) of Schedule 1AB to TMA is satisfied and the Appellants have overpaid tax but it is HMRC who are asserting that the SATRs were filed in accordance with the PGP at the date of filing; it is therefore HMRC that are required to satisfy the burden of proof.[58]In Rose, Finlay J in the High Court confirmed that the question of whether there was a PGP was a question of fact. The appeal by way of case stated was against the decision of the Special Commissioners, who had heard evidence, were satisfied that the practice generally prevailing in the case of hire purchase agreements had been to deduct the average and not the actuarial hire of railway wagons. Finlay J, in dismissing the appeal, stated: “… the first thing to ascertain – if any point arises with regard to it- whether there has been this practice. That is a pure question of fact and a question of fact which is, as I read it, to be determined by the Commissioners.”.[59]We note that HMRC’s own guidance in the Self Assessment Claims Manual at SACM12105 headed “Overpayment relief: Exclusions: Cases G and H – Practice generally prevailing accords with our view and relevantly states (emphasis added):
“Case G - Capital gains and income tax other than PAYE income Overpayment relief is not due if the claim relates to a mistake in an SA return or other tax calculation and the tax liability was calculated in accordance with the practice generally prevailing, see below, at that time, except where the claim relates to PAYE income. Practice generally prevailing Whether there was a “practice generally prevailing” is a question of fact (Rose Smith 17 TC 586). In HMRC v Household Estate Agents Ltd Henderson J. stated ‘Without attempting to give an exhaustive definition, 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.’ In relation to overpayment relief, the onus is on HMRC in any appeal hearing to demonstrate that there was a practice generally prevailing. You may need to refer, among other things, to our published guidance, advice from HMRC technical specialists, reported cases and external comment as evidence of a practice generally prevailing. A practice need not have been universally followed. But where a tribunal or court decides that a practice is wrong you should take it to have ceased to be a generally prevailing practice at that point even if the decision is subject to appeal.”
Practice generally prevailing ‘Without attempting to give an exhaustive definition, 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.’

Relevant date of PGP

[60]HMRC submitted that when considering whether there was a PGP, the relevant date is the date that the SATR was filed, s29(2) TMA. Whilst we agree with Mr Good’s submission that s29(2) TMA is concerned with discovery assessments we consider that the relevant date for considering whether the SATR was made in accordance with the PGP can only be the date that the SATR was filed. We consider that is clear from the wording of paragraph 2(8)(b) of Schedule 1AB to TMA- “liability was calculated in accordance with the practice generally prevailing at the time”.

Evidence of PGP

[61]Both parties relied upon Boyer at [34] to [40] which identified the requirements for a practice to be considered PGP. In Boyer, Judge Berner, considered Rose, Rafferty v R & C Commrs (2005) Sp C 475 (“Rafferty”) and the three elements identified by Henderson J in Household at [58]. Judge Berner then set out the features that he identified as relevant for a practice to be considered PGP. We agree with and adopt the requirements identified by Judge Berner at [34] to [42] which we have summarised below:(1) A practice must be capable of being readily ascertained; the practice must have substance and be sufficiently precise and devoid of uncertainty as to its application [34];(2) A practice will be readily ascertainable not only through published statements but also if it was possible for something to be identified as the practice if it were settled, defined and agreed between HMRC and taxpayers, or communicated between HMRC and taxpayers, or otherwise sufficiently identified to the outside world [35];(3) The requirement for a practice to be relatively long-standing emphasises that such practice is not confined to published statements but extends to unpublished practices. A published practice is likely to be capable of being regarded as “generally prevailing” over a shorter period than one that merely becomes established through practice [36];(4) An internal practice adopted by HMRC will not be generally prevailing until such time as it can be identified with sufficient precision by taxpayers and their advisers [37];(5) The same quality of clarity and precision must be present in the understanding of HMRC and taxpayers and their advisers alike [38];(6) In order that a practice may be regarded as generally prevailing, it must have been adopted by HMRC and generally, if not universally, by the taxpayer community [39];(7) The practice must be settled and applied in a consistent manner [40];[62]We have considered those requirements in turn. A practice must be capable of being readily ascertained; the practice must have substance and be sufficiently precise and devoid of uncertainty as to its application[63]We consider that the practice was readily ascertainable, had substance and was precise as set out in IPTM3820 and IPTM3840 both published on 5 August 2016. IPTM3840 sets out clearly and precisely seven steps that may be required to be applied to the gain in order to calculate the amount of TSR. Depending upon the taxpayer’s circumstances, not all seven steps may need to be applied but that position is made clear. Mr Good relied upon Judge Berner’s reference to inchoate at [34] in Boyer and referred to the dictionary definition of inchoate of which the most appropriate synonyms are: basic; crude; elementary; rough and ready; and undeveloped. At [34],

Judge Berner stated:

“the practice must have substance (in the sense of not being inchoate)”
. The OED defines “inchoate” as “Just begun, incipient; in an initial or early stage; hence elementary, imperfect, undeveloped, immature.”. We consider that this definition is to be preferred in the context of contrasting “inchoate” with “substantial” and that the practice identified in IPTM3820 and IPTM3840 had substance. It was accepted by both parties that the seven steps for calculating TSR were embedded in the Self Assessment calculator and commercially available tax calculation software. Therefore, there could be no uncertainty as to its application despite it subsequently being found to be incorrect. A practice will be readily ascertainable not only through published statements but also if it was possible for something to be identified as the practice if it were settled, defined and agreed between HMRC and taxpayers, or communicated between HMRC and taxpayers, or otherwise sufficiently identified to the outside world[64]We consider that the requirements are met. Both IPTM3820 and IPTM3840 are available in HMRC’s Insurance Policyholder Taxation Manual which is published on the gov.uk website and are publicly accessible. The practice has been sufficiently identified to the outside world. The requirement for a practice to be relatively long-standing emphasises that such practice is not confined to published statements but extends to unpublished practices. A published practice is likely to be capable of being regarded as “generally prevailing” over a shorter period than one that merely becomes established through practice[65]We consider that the requirements are met. We did not understand this element to be disputed by Mr Good. In any event, both IPTM3820 and IPTM3840 are a “published practice” Both IPTM3820 and IPTM3840 were updated on 5 August 2016. We were not provided with copies of the previous versions but are of the view that the reference to “updated” confirms that the practice was relatively long-standing. An internal practice adopted by HMRC will not be generally prevailing until such time as it can be identified with sufficient precision by taxpayers and their advisers[66]This requirement was not relevant to the issues that we are required to determine. The same quality of clarity and precision must be present in the understanding of HMRC and taxpayers and their advisers alike[67]We do not consider that this requirement is met. The correspondence relied upon by Mr Good confirms that the same answer to an enquiry about TSR would not be given by the broad community of tax professionals engaged in this particular area. This point is considered further at paragraph 93 below. In order that a practice may be regarded as generally prevailing, it must have been adopted by HMRC and generally, if not universally, by the taxpayer community[68]We do not consider that this requirement is met. HMRC relied upon the following as evidencing the existence of a PGP when the Appellants’ filed their SATRs: facts from the decision in Silver, the Appellants’ SATRs as originally filed, guidance contained in Insurance Policyholder Taxation Manual at IPTM3820 and IPTM3840 both dated 5 August 2016, absence of any tax disputes or any Tribunal decisions where TSR was challenged.[69]HMRC submitted that the fact that the Appellant in Silver was unrepresented by an accountant, solicitor or similar tax professional (“Representative”) and had completed the tax return and TSR computation herself without the assistance of a Representative confirmed the existence of a PGP. The Appellant in Silver was unrepresented as Representatives did not agree with the Appellant’s arguments. We do not accept HMRC’s assertion. There is no evidence before the Tribunal nor recorded in the decision that Mrs Marina Silver was represented by her husband, Mr N Silver, because no Representative was willing to act because the Grounds of Appeal were contrary to a settled PGP. As HMRC are aware, there a variety of reasons why a taxpayer is unrepresented, the main reasons are often the cost of instructing a Representative relative to the amount of tax at stake (£21,391) or the disputed matter is considered straightforward and capable of being dealt with by the taxpayer. Mr Good had put in evidence what he described as Mr Silver’s witness statement but were Mr Silver’s written note of his submissions. It is clear from Mr Silver’s written submissions that he considered the matter straightforward and within his capabilities. The introduction at paragraph 1 of the verbal statement confirmed that Mr Silver did not “have any legal or accountancy qualification or experience but have attempted to research the legal background of this particular tax demand in what I believe to be an intelligent and logical manner … I am a Chartered Civil Engineer by profession and, although at the age of 87 … I am still capable of understanding mathematical formulae and their intentions”. It is clear from the correspondence between HMRC and Representatives, considered below at 93, that a broad body of Representatives did not adopt HMRC’s interpretation of the legislation and its methodology for calculating TSR.[70]HMRC relied upon the fact that Mrs Silver was charged with a penalty for careless inaccuracy as HMRC had concluded that her return was inaccurate and not in accordance with the PGP at that time. HMRC agreed to suspend the penalties and there was no appeal against them. We can accept that a penalty for careless inaccuracy was imposed as the return was inaccurate based on HMRC’s view of the legislation but do not accept that imposing penalties evidences that the return was not completed in accordance with the PGP at that time. As stated in Boyer at [39], in order that a practice may be regarded as generally prevailing, it must have been adopted by HMRC and generally, if not universally, by the taxpayer community. The fact that HMRC have adopted such a practice does not, without the general if not universal adoption by the taxpayer community, make it a PGP.[71]HMRC additionally relied upon the fact that they had appealed Silver to the UT (but withdrew after new legislation was announced) as evidencing the existence of the PGP. We do not accept that appealing the Silver decision to the UT evidences the existence of a PGP, all that appealing the decision to the UT confirms is that HMRC considered that Silver contained an error of law. In Sally Judges at [56]

Judge Brown KC stated:

“56. HMRC litigated Silver on the basis that the hypothetical recalculation did not permit account to be taken of a personal allowance which had been reduced as a consequence of the chargeable event gain and they lost. An appeal must have been lodged on the basis that Judge Mosedale had made an error of law. But it was subsequently withdrawn in close proximity to the making of a legislative amendment. Therefore there must be a strong indication that the judgment was not wrong but that ministers were concerned that the interpretation adopted would carry consequences which had not been addressed in Silver. Legislation was therefore introduced and ultimately enacted to address those consequences of which the facts of this case are an illustration. It is somewhat difficult to conclude that applying the legal principles determined by Judge Mosedale to the facts of this case the factual differences would impact her decision a full section 23 ITA calculation is required. A full calculation s23 ITA calculation is a full section 23 ITA calculation whether HMRC like the outcome or not.”
[72]The analysis of HMRC’s TCG relied upon by the Appellants was not challenged by Mr Asuelimen and we accept that the providers of tax computational software do not deviate from HMRC's computation methodology except in very rare circumstances. We do not accept HMRC’s submission that the fact that the Appellants’ TSR was calculated in accordance with HMRC’s interpretation leads to the conclusion that there existed a PGP at the time of the Appellants filing their SATRs. The requirement identified in Boyer is that in order for a practice to be a PGP “it must have been adopted by HMRC and generally, if not universally, by the taxpayer community”. The correspondence and documents relied upon by Mr Good (considered below) confirm that the absence of universal acceptance and general acceptance of HMRC’s TSR calculation methodology by the taxpayer community; the fact that the Appellants’ returns applied HMRC’s TSR calculation methodology embedded in commercially available software is not fatal to their OPR claim as the existence of the PGP has to be determined by reference to the overall taxpayer community and not by reference to individual agents or taxpayers.

The practice must be settled and applied in a consistent manner

[73]We do not accept that this requirement has been met. In our judgment, that conclusion is clear from the Appellant’s evidence considered below at X. HMRC relied upon the following documents as evidencing the existence of the PGP until 11 March 2020: Canada Life document titled “Calculating top-slice relief on a chargeable gain Briefing Note (undated), FT Adviser article titled “Advisers must understand how to calculate top slicing relief” dated 20 May 2020, Pru Adviser article titled “Top slicing relief for bonds taxation: the facts” stated to be last updated on 27 April 2022, Abrdn Techzone article titled “Top Slicing Relief” dated 6 April 2022 and Tolley article titled "FA 2020 – changes to top slicing relief” dated 7 August 2020.[74]We note that the Canada Life briefing note is undated but that reference is made at the bottom of the first page it is stated “As of 11 March 2020, HMRC confirmed that for the top-slicing calculation only, the personal allowance should be reinstated …” confirming, at the very least that it was published after 11 March 2020. We agree with Mr Good that this article is of limited evidential value as it post-dates the decision in Silver and the Chancellor’s Budget speech on 11 March 2020, it does not provide any evidence of the PGP before that date.[75]The FT Adviser article dated 20 May 2020 again post-dates the Chancellor’s Budget speech on 11 March 2020. The article explains what TSR is, outlines the changes HMRC has made to calculating TSR and explains how allowances are determined. The article states “Top slicing relief has been a hot topic for the past year or so. Since the beginning of 2019 we have seen HM Revenue & Customs change its calculation method, lost a First-Tier Tribunal tax case (the Silver case), introduce new legislation in the Budget and withdraw its appeal against the decision in the Silver case. This has kept the calculation of top slicing relief firmly in the news. … The key thing to understand is that the shorthand method of top slicing many advisers will have relied upon to work out the additional tax on the bond gain will no longer be reliable in many cases This means advisers need to get to grips with the new calculation method to be able to accurately determine the tax payable when clients have bond gains. … Relief is only available if some part of the full gain is subject to tax at higher or additional rate when added to other income. … If any part of the gain is taxable at a higher rate, top slicing relief will be available as reduction to the overall tax bill …The 2020 Budget provided additional clarity on the availability of personal allowance for the purposes of calculating top slicing relief on gains made on or after March 11 2020. … It is therefore crucial that advisers fully understand how top slicing relief is calculated and applied and do not rely upon the shorthand method”[76]Mr Good submitted, and we agree, that what the article does show, in contradiction to HMRC’s submission, is support for the Appellants’ position that advisers have long had difficulty in ascertaining the correct practice.[77]The Pru Adviser article dated 27 April 2022 explains what taxpayers need to know about TSR and how it works. The article states under the sub-heading of Personal Savings Allowances: The Personal Savings Allowance and the Starting Rate for Savings The starting rate for savings and the personal savings allowance nil rate should be taken into account when calculating top slicing relief, where applicable. The starting rate for savings is available to those taxpayers with total non-savings income of less than their personal allowance plus £5,000. If taxable earned and other non-savings, non dividend income is above the £17,570 2022/23 (or £20,170 for those eligible for the blind person’s allowance) , the starting rate for savings will not apply to the taxable savings income. The personal savings allowance nil rate band is applied to the first £1,000 of savings income for basic rate taxpayers. and the first £500 for higher rate taxpayers. The personal savings allowance and the starting rate for savings are nil rate tax bands and are therefore ‘allowances’. Despite that, in IPTM3820 HMRC state: “These allowances are not adjusted when calculating the notional tax due on the ‘sliced gain ’” … In this tax case [the hyperlink is to the decision in Sally Judges] relating to the 2017/1818 tax year, the First-tier Tribunal confirmed that the legislation introduced in 2020 removing beneficial ordering was not clarification of existing legislation and did not have retrospective effect. The decision was released on the 18 February 2022 and is within the 56-day appeal period for HMRC to challenge. Tim Good (of Absolute Accounting Software Limited) who represented the taxpayer in the above case commented that the judge agreed that when applying beneficial ordering, legislation requires that the income tax liability rather than income tax payable is to be minimised (i.e. before rather than after deducting any notional bond tax credit). This may occasionally result in higher overall tax payable than calculated by the current HMRC self-assessment calculator in respect of gains arising on or after the 11th of March 2020.[78]Mr Good again submitted, and we agree, that what the article does show, in contradiction to HMRC’s submission, is support for the Appellants’ position that advisers have long had difficulty in ascertaining the correct practice to be applied when calculating TSR.[79]The abrdn techzone article dated 22 April 2022 explains what is top slicing relief, explains the calculations and provides 3 example scenarios. We note that under the heading of related content is a hyperlink to an article titled “Top slicing relief – could the shorthand method leave your clients short changed?” Mr Good submitted that second scenario in the abrdn techzone article incorrectly calculated the TSR as £2,100 when the correct TSR figure following the changes made in the FA 2020 is £2,340. This he submitted showed the continuing difficulties that advisers and insurance providers faced when ascertaining the correct way to calculate TSR. Again, the article post-dates the post-dates the Chancellor’s Budget speech on 11 March 2020 and we do not consider that the article provides any evidence as to what was the PGP at the time the Appellants filed their SATRs.[80]The Tolley article date 7 August 2020 is a news item that discusses the changes made by s37 FA 2020 and its application to individuals and the position regarding chargeable event gains arising prior to the change in the legislation. The articles states under the heading of “Overview of chargeable event gains and top slicing relief” that “the policyholder is only liable to extra tax on life insurance gains if they are a higher rate or additional rate taxpayer. We agree with Mr Good that the article confirms the continuing confusion as to the correct basis for calculating TSR. The article notes the treatment of the savings nil-rate band and questions HMRC’s interpretation of the legislation and suggests that, following the Tribunal decision in Silver, a taxpayer who wished to litigate the savings nil-rate band point would have a good prospect of success but that the amount of tax at stake would make any litigation not cost effective:
“What if the chargeable event gain takes the taxpayer into a different tax band, thus reducing the amount of their savings nil rate band? HMRC’s view is that the amount of the savings nil rate band is determined by the taxpayer’s taxable income including the full chargeable event gain. Therefore, even if the annual equivalent of the chargeable event gain would mean that they would fall into a lower tax band and should be due a larger savings nil rate band, top slicing relief must be calculated based on the savings nil rate band available (if any) on their taxable income including the full chargeable event gain. See Example 4 in IPTM3820. In light of the Silver decision regarding the taxpayer’s entitlement to the personal allowance, this seems surprising. If Parliament’s intention was for top slicing relief to be determined based on the entitlement to the personal allowance based on the annual equivalent, would this not extend to the amount of the savings nil rate band? Silver v HMRC [2019] UKFTT 263 (TC) The Silver decision plus the fact that the Government decided to amend the law to reflect this rather than appeal the decision would suggest that any taxpayer who wished to litigate the savings nil rate band point would have a good chance of success. However, as the amount of tax involved is not likely to be very high, it may not be cost effective for the taxpayer to take this to the Tribunal.”
We do not consider that the article supports HMRC’s submission.[81]In our judgment, the documents and evidence relied upon by HMRC do not provide evidence of a PGP at the date that the Appellants’ submitted their SATRs and therefore HMRC have not discharged the burden of proof in establishing a PGP. The extensive evidence relied upon by Mr Good supports that conclusion.[82]Mr Asuelimen submitted that the publication of one article and correspondence with one individual, whilst raising questions, is not in and of itself sufficient to demonstrate an entire or wholesale departure of the community of tax advisers from a PGP that had been in operation for a considerable period of time. We do not accept that submission. The evidence relied upon by Mr Good is considered below.

Published articles

[83]The first article questioning HMRC calculation methodology was published in an article by Richard Curtis titled “L.A. Shakedown” in Taxation on 31 July 2002 ( Issue 3868). The article considered at length s550 Taxes Act 1988 and questioned whether the “’top-slicing’ provisions of section 550 only apply to higher rate liabilities? Although the title mentions it, I can find no mention of higher rate tax in the section itself”. The article concludes stating:
“Unfortunately, the background notes in respect of clause 86 contain the phrase (and let us hope that this does not comprise 'financial advice', otherwise taxpayers may start taking Mr Marris's [a Labour MP taking part in Standing Committee Fs review of the Finance Bill 2002] the suggestion to heart) that, 'There is no further tax charge when a policyholder is not liable to tax at the higher rate'. “If the Revenue can get it wrong, what hope is there for Mr Mann [the taxpayer in Mr Curtis’ worked calculations]?”
We agree with HMRC that the publication of this comment piece did not change the PGP at the date of publication.[84]A further article by Richard Curtis titled ”The heat is on” with the sub-heading “Seven years have passed, and RICHARD CURTIS still thinks that basic rate pensioners should be entitled to top-slicing relief” was published in Taxation on 8 September 2009 (Issues 4222). The article analysed the changes made as part of the Tax Law Rewrite project and the introduction of ITTOIA 2005. Mr Curtis set out his view that HMRC’s interpretation of the top-slicing relief calculation is flawed and concluded stating: “I wonder if it is not time for HMRC to adopt a new interpretation of s 535 et seq. Presumably there would be no historic relief as those calculations would be 'in accordance with prevailing practice'. Am I onto something here or would that still just be a 'novel interpretation of the legislation'? In our view, the final two paragraphs confirm that HMRC’s interpretation of “s535 et seq” was the practice generally prevailing at the time of the article and what Mr Curtis was positing was possibly a “novel interpretation of the legislation” contrary to the PGP at that time.[85]On 28 September 2017, an article by Mr Tim Good titled “It’s all gone Pete Tong” was published in Taxation (“the Article”). The sub-heading stated “TIM GOOD believes that that HMRC and tax practitioners have been calculating top-slicing relief incorrectly. He suggests that it is time to follow the legislation rather than existing practice”. We note the reference to “existing practice” which, in our view, confirms the existence of the PGP at the date the Article was published. The Article set out in considerable detail and by reference to the legislation what Mr Good believed was the correct calculation for TSR which was subsequently confirmed as correct in the Tribunal decisions in Silver and Sally Judges. The Article’s conclusion stated:
“Conclusion So what should tax practitioners do in such cases now? My main recommendations are as follows. First, do not rely on the HMRC calculation of top slicing relief. Second, do not rely on your tax return software – almost all the main brands (my own included) clone the HMRC calculation. Third, review all clients who have had chargeable event gains in recent years and file amended returns if possible. Finally, consider an appeal to the tribunal if HMRC refuses to accept that its calculation is wrong.”
[86]In our judgment, the publication date of the Article, 28 September 2017, is the date that the PGP for calculating TSR ended. We have reached that conclusion following consideration of the correspondence at X below, the correspondence confirms the Article clearly identified the errors in HMRC’s TSR computational methodology and was the catalyst for the “taxpayer community” challenging HMRC’s practice.[87]The Financial Times Adviser article dated 14 February 2019 titled “HMRC admits tax calculator was wrong” reiterated the view set out by Mr Good in the Article and refers to the confusion around TSR by professional insurance providers. The article stated:
“HM Revenue & Customs has fixed its self assessment tax calculator after the tax office was told it was incorrectly stating tax bills for withdrawals from life policies. Top slicing relief applies to life policies and means if a client had a life policy which was held for 10 years, and paid out £40,000, then the cash amount is divided by the number of years held, to create a one-year value. The tax band that this sum sits within is the band at which tax is paid for the whole amount. Tim Good, a tax specialist, said he met with HMRC to highlight how, since changes were made to the personal tax allowance in 2010, the nation's tax office has been incorrectly calculating the tax liability, by excluding some higher rate taxpayers from the tax-free allowance to which they are entitled. Instead of top slicing relief being properly applied, Mr Good said individuals were being taxed on their income from the policy for the year in which it was taken - so in some cases, basic rate taxpayers have been taxed as higher rate taxpayers. … Mr Good urged advisers to challenge decisions from HMRC on the issue of tax liability on life policies. He said: "My view is their calculator is wrong. They say it isn't. The decisions are being challenged and it is likely that there will be tax tribunals on this in the coming months." But a representative of HMRC said the tax office had been aware of an issue with the calculator and this had now been fixed. He said: "We have corrected our self assessment calculator to ensure top slicing relief is calculated correctly." Speaking at FTAdviser's Tax Efficient Investing event in London, Neil Jones, market development manager of iCan at Canada Life, said HMRC did need to clarify their approach. He said: "There is a lot of confusion around the top slicing allowance. The Association of British Insurers has been working with HMRC around this guidance. HMRC have got a lot of things on their plate at the moment but when we get guidance we will pass that on.”
[88]On 23 May 2019, Taxation published an article by Mr Good titled “Silver wins gold” which reviewed the decision in Silver and, having set out the requirements for overpayment relief claims, advised agents to submit overpayment relief claims to HMRC.

E-mail correspondence between HMRC and Mr Good

[89]The e-mails relied upon by Mr Good cover the period 11 October 2017 to 30 July 2018. The initial e-mail from Mr Good to HMRC dated 11 October 2017 was sent to named individuals whose e-mail addresses had been provided to Mr Good by the ICAEW Tax Faculty as the appropriate individuals to contact to discuss the computation of TSR. The e-mail stated:
“The ICAEW Tax Faculty have given me your email address (and those of Rosa and David) as the HMRC contacts with whom to discuss the computation of top slicing relief on chargeable event gains. By way of background, I have had various discussions with your colleagues at SDST (especially Keith Graham and Tony Musk) concerning the 2016-17 income tax calculation and the exclusions. I think that some of my algorithms are now embedded in the HMRC system. Over the summer I have developed further algorithms to calculate top slicing relief and concluded that the HMRC calculation takes a simplified approach which, although correct in many cases, significantly understates TSR in a number of cases (in some by tens of thousands of pounds) and occasionally overstates TSR in others (although not by more than £1,000). As far as I am aware, all the commercial developers have copied the HMRC methodology in their tax return software. I attach a copy an article by me, published last month in Taxation Magazine, which sets out my calculations and gives examples. The article, together with my lecture presentations (my day job is lecturing to accountants on the CPD circuit), is generating considerable interest amongst agents and I think that HMRC will soon experience quite a high level of demand as a result. It would be extremely helpful if somebody in the Department could "take ownership" of this issue with a view to giving an official response and to that end I would be grateful for a response. I would be more than happy to have a meeting to explore this issue further.”
I would be more than happy to have a meeting to explore this issue further.”[90]The reply, of the same date, from Mr Darryl Wall acknowledged safe receipt and confirmed:
“We will get back to you once I have had time to fully consider the points you have raised.”
The subsequent e-mail exchanges dated 6 February 2018, 18 June 2018, 13 July 2018 and 30 July 2018 between Mr Good and various individuals at HMRC dealing with TSR confirm HMRC’s engagement with Mr Good, acknowledgement that HMRC’s SA calculator did not in some instances calculate TSR correctly, identification of an error in the SA calculator and that Mr Good had obtained Counsel’s opinion which agreed with Mr Good’s analysis in the Article.

AccountingWEB postings

[91]Mr Good provided five “chat” postings from the AccountingWeb site referring to TSR issues. The “about AccountingWEB” section on the AccountingWEB site states:
“AccountingWEB.co.uk is the largest independent online community for accounting and finance professionals in the UK - providing award-winning content and online engagement between members in a true community environment. Accountants in Practice represent just under half of our visitors and we reach 75% of firms outside of the top 20. Accountants in Business represent the remainder of our audience, typically working in smaller to mid-tier businesses.”
[92]The “chat” postings (started by different individuals) all refer to TSR with requests for advice and/or recount their experience of dealing with HMRC are dated 20 March 2018, 28 May 2019, 19 August 2019, 12 March 2020 and 23 September 2022. The posting dated 20 March 2018 refers to using “the calculations in Tim Good’s Taxation article of September 2017, top slicing relief would be due but I am aware that HMRC have rejected this analysis and it appears that no relief would be due using their calculation”. One posted reply refers to similarly querying HMRC’s calculation “following a course run by Tim Good” with another posted reply requesting advice on calculating TSR. The posting dated 12 March 2020 with 14 replies contained a reply timed at 15.28 which stated:
“this has been a simmering gripe for a few years now - it essentially started after a raft of exclusions were published a week or so before January 2018”
. The posting dated 23 September 2022 contains 17 replies which confirmed the continuing confusion amongst AccountingWEB members on how to correctly calculate TSR.

Correspondence with taxpayer agents

[93]Mr Good (with permission) provided over 50 examples of copy correspondence between taxpayer agents and HMRC in which the agents challenged or disagreed with HMRC’s calculation of TSR. The copy correspondence spans the period 12 October 2017 to 19 October 2022; however, reference is made in the correspondence to disputes pre-dating 12 October 2017. Reference is made in an e-mail from Tim Good to Mr Terry Lloyd CTA dated 14 August 2019 “This is the first in what might become a tedious serious of emails that I am sending to all those (about 150) who have contacted me personally about top slicing relief on chargeable event gains.” In the interests of brevity, we have not referred to each and every correspondence chain provided to Mr Good but have merely referred to the selected correspondence below. We are satisfied that correspondence in evidence before the Tribunal confirms the following:(1) The agents are located across the UK;(2) The agents vary in size from sole practitioners, High Street practices; mid-level firms with several offices and large national organisations such as the Co-op Legal Services and Deloitte LLP;(3) In various instances HMRC accepted the agents’ TSR calculations. The agents and taxpayers’ challenges to the HMRC’s TSR calculations post-dated Mr Good’s Taxation article dated 28 September 2017 but pre-dated the Silver decision: (a) Letter from Advance Accountancy to HMRC dated 4 December 2018:
“the top slicing relief is not calculated correctly”. (b) Letter from HMRC dated 14 September 2018 to Co-Op Legal Services stating: “Thank you for your telephone call of 13 September 2018, querying the amount of Top Slicing Relief”. (c) Letter from Willow Accountancy dated 13 March 2019 to HMRC referring to correspondence in 2018 challenging HMRC’s TSR calculation. (d) HMRC letter to Harwoods dated 23 May 2019 referring to an OPR claim dated 25 February 2019. (e) E-mail from Deacons Accountants to Tim Good dated 28 June 2022 confirming that HMRC’s TSR calculation was challenged on 16 October 2017. (f) E-mail from Teddington Tax Services Ltd dated 7 January 2021 referring to appeal against TSR made on 26 January 2018. (g) Letter from Carter Rose Associates to HMRC dated 3 December 2018 challenging the TSR calculation. (h) Letter from HMRC to MA Partners LLP dated 20 September 2019 referring to agent’s letter dated 14 September 2018 challenging HMRC’s TSR calculation. (i) Letter from Mr PM Peters (Executor and Tax Partner, Whiting and Partners) to HMRC dated 8 January 2018 disputing HMRC’s TSR calculation. (j) Letter from Chipchase Manners to HMRC dated 18 February 2019 referring to 2018 tax return filed on 28 January 2019 attaching an explanation of how the agent had calculated TSR and challenging HMRC’s incorrect calculation. (k) letter from Rushton Accountants to HMRC dated 21 November 2018 challenging TSR calculation. (4) In some instances, paper SA returns were filed to avoid HMRC’s SA Calculator providing an incorrect calculation of TSR: (a) HMRC letter to Edwards Pearson & White dated 16 October 2019 acknowledging paper return for ye 5 April 2018. (b) Letter from Pope & Co to HMRC dated 1 January 2018. (c) Letter from Tony R Pomfret to HMRC dated 15 November 2018 disputing HMRC’s TSR calculation stating that was the reason why the paper return was filed. (d) Letter from Murphy Salisbury to HMRC dated 5 September 2018 enclosed paper SA return as the client had a chargeable gain. (e) E-mail from Mercer Lewin to Tim Good dated 9 January 2019 confirming that a paper SA return sent to HMRC with reasons for the agent’s TSR calculation. (f) Letter from HMRC to Carson & Trotter acknowledging letter dated 9 November 2017 challenging TSR calculation. (g) Letter from Wildes Chartered Accountants to HMRC dated 11 September 2018 challenging TSR calculation. (h) Mac Kotecha & Co submitted a paper SA return to HMRC on 19 June 2018. (5) References are made in correspondence to relying upon Tim Good’s TSR calculator for the correct calculation when completing the return: (a) Co-op Legal Services e-mail dated 2 May 2019. (b) e-mail from Shaw Gibbs dated 8 February 2021 re 2018 TSR calculation. (c) Mac Kotecha & Co submitted paper return letter on 19 June 2019 to HMRC using Tim Good’s TSR calculator. (6) Reference is made to agents becoming aware of the TSR calculation issue following publication of the Article or following a professional training event: (a) E-mail from Deacons Accountants to Tim Good dated 28 June 2022: “My own dealings with TSR started with a challenge dated 16 October 2017, following my reading of your 28 September 2017 article.”. (b) E-mail from On the Spot Accountants to Tim Good dated 19 October 2018: “I have a client caught up in this issue on his 2017 tax return. I’ve argued since January that HMRC’s software is wrong. This was based on your Taxation Article and from attending PTOP Updates.” (c) E-mail from Mr PM Peters (Executor and Tax Partner Whiting and Partners) to Tim Good dated 21 February 2019 stating: “I wonder if there has been any progress on this matter from your perspective since you first brought it to our attention in September 2017?”. (d) E-mail from Terry Lloyd CTA to Tim Good dated 14 August 2019 “After your articles I resubmitted the calculations” [1 February 2018]. (e) E-mail from Abacus Network to Tim Good dated 19 September 2018 stating “I read with great interest you article in Taxation last September and your method of calculating TSR that you recommended has a big impact on my client’s tax liability” and an e-mail dated 31 October 2018 “This week I filed an appeal against HMRC’s calculation of Top Slicing Relief for 2017-18”
HMRC subsequently withdrew from the appeal.” (f) E-mail dated 12 October 2017 from HWB Chartered Accountants to Tim Good stating “I have, with great interest, watched the TaxTV episodes from both September and October 2017 and read your article in Taxation with regard to chargeable events and the way in which top slicing relief is incorrectly calculated by HMRC”. (7) In two instances, HMRC confirmed they would not be defending appeals before the Tribunal. One withdrawal dated 19 October 2022 confirmed that HMRC had “reviewed the grounds and calculations put forward by the Appellant’s representative and are prepared to accept them in this specific case”. (8) HMRC in some instances acknowledged that the SA Calculator was inaccurate: (a) Letter dated 19 April 2018 from HMRC to Deacon’s Chartered Accountants stating the SA Calculator “hasn’t been producing accurate calculations in some instances when considering [TSR] on chargeable event gains”. (b) Letter from HMRC to An Accountant & Co dated 24 November 2020 in respect of 2016/17 return submitted pre-December 2018:
“There have been a few ongoing issues with the self-assessment calculator following the introduction of the new nil savings rates in April 2017 … This meant our software did not initially calculate the right amount of tax due.”. (c) HMRC letter to Tony R Pomfret & Associates dated 20 December 2019: “there have been a number of ongoing issues with the self-assessment calculator following the introduction of the new nil savings rates in April 2016”. (9) HMRC acknowledged that it was aware that Representatives were challenging TSR calculations: (a) HMRC letter to Co-op legal services dated 1 May 2019: “HMRC is well aware that accountants and agents are querying TSR calculations issued by HMRC”
.

Conclusion

[94]In our judgment, HMRC have not satisfied the burden of proof in establishing the existence of a PGP at the time that the Appellants filed their SATRs. That conclusion is sufficient to allow the appeals; however, the evidence and documents relied upon by the Appellants supports that conclusion and that the PGP ceased to be the PGP on 17 September 2017 when the Article was published.

Decision

[95]For the reasons set out above, we allow the Appellants’ appeals against the closure notices issued on 18 June 2021(in respect of Mr Joye) and 1 November 2021(in respect of Mr Sumner) and the respective overpayment relief claims submitted on 5 May 2020 (Mr Joye in the sum of £55,201.63) and 25 March 2021 (Mr Sumners in the sum of £40,758) are given effect.

Right to apply for permission to appeal

[96]This document contains full findings of fact and reasons for the decision. Any party dissatisfied with this decision has a right to apply for permission to appeal against it pursuant to Rule 39 of the Tribunal Procedure (First-tier Tribunal) (Tax Chamber) Rules 2009. The application must be received by this Tribunal not later than 56 days after this decision is sent to that party. The parties are referred to “Guidance to accompany a Decision from the First-tier Tribunal (Tax Chamber)” which accompanies and forms part of this decision notice. Release date: 05th JUNE 2025