Pulak Rakshit & Ors v The Commissioners for HMRC

I will begin by describing the steps undertaken by the appellants as part of the scheme. There is no material dispute between the parties as to these facts. Each of the appellants undertook these transactions in substantially the same way. All transactions took place in 2010 or 2011.

The steps in the scheme were as follows.

SDLT is usually payable on the “chargeable consideration” for a “land transaction” involving a transfer of a “chargeable interest” in land. A freehold or leasehold is a “chargeable interest” and so, if they had not taken the steps involved in the scheme, the appellants would, in the normal course, have expected to pay SDLT on their purchases of the properties.

The intended effect of the steps in the scheme was that no SDLT would be payable on those purchases. This was on the basis that:

As I have described, the transactions in the scheme were based on the transactions involved in the Project Blue case. The appellant(s) withdrew their arguments that there was no liability to SDLT on their transactions following the decision of the Supreme Court in the Project Blue case that section 75A applied to impose SDLT on the acquisition of the property in that case.

I will refer to the decision of the Supreme Court in the Project Blue case in more detail later in this decision notice, but, before I turn to the facts surrounding HMRC’s enquiry into the implementation of the scheme by each of the appellants, I will first summarize the facts of and decisions of the courts and tribunals in the Project Blue case as it provides some useful background to the progress of those enquiries.

The Project Blue case involved the purchase by a company, Project Blue Limited (“PBL”), of the site of the Chelsea Army Barracks from the Ministry of Defence for £959 million. PBL financed the acquisition of the site through an Islamic financing arrangement with a Qatari bank. Under that arrangement, PBL agreed to sub-sell the freehold of the site to the bank and the bank agreed to lease the site back to the company. The company and the bank also entered into various options which entitled PBL in due course to re-acquire the freehold from the bank.

PBL filed a land transaction return, claiming that no liability to SDLT arose on its purchase of the barracks from the Ministry of Defence because sub-sale relief under section 45(3) FA 2003 applied. The bank filed a return in relation to the sale of the freehold to it by PBL, claiming relief under section 71A FA 2003 (“alternative property finance”). PBL also filed a return in relation to the grant of the lease by the bank to the company. The company claimed relief under section 71A FA 2003 in relation to the grant of the lease.

In a decision issued on 18 December 2014, the Upper Tribunal found that the transactions did not attract SDLT but for the application of the anti-avoidance rule in section 75A FA 2003. However, SDLT was payable by PBL under section 75A FA 2003 on the purchase price of £959 million.

The Court of Appeal overturned that decision in a judgment issued on 26 May 2016. The Court of Appeal found that section 45(3) applied to exclude the transfer of the barracks to PBL from SDLT, but that section 71A(2) did not exempt the bank from liability on the transfer of the barracks to it under the financing arrangements. This was on the basis that PBL could not be the “vendor” for the purpose of section 71A(2) as the contract between the Ministry of Defence MoD and PBL for the purchase of the barracks was disregarded under section 45(3) FA 2003.

The Supreme Court released its decision on 30 June 2018. The majority found that section 45(3) and section 71A applied to exempt the relevant transactions from SDLT. However, PBL was liable to SDLT under section 75A FA 2003 on a notional transaction between the Ministry of Defence and PBL.

My findings of fact are set out in this section.

The parties conducted their arguments by reference to the transactions undertaken by the first appellants, Mr Pulak Rakshit and Mrs Sharmilla Rakshit. I will also describe the steps in the scheme by reference to the transactions undertaken by them. Unless otherwise stated, these findings apply to all of the appellants.

The first appellants appointed ELS International Lawyers LLP (“ELS”) as their agent. (Each of the other appellants also appointed ELS.)

The first appellants agreed to purchase a property at 14a Tangier Road, Guildford from Ms Gillian Botwright (the “original vendor”) for £656,000 under a contract dated 20 August 2010 (the “original contract”).

After the exchange of contracts but before completion, the first appellants entered into an agreement with Vale to transfer the property to Vale at the same time as the completion of the original contract for £656,000 under an alternative finance arrangement, under which Vale also agreed to grant a 999-year lease of the property back to the first appellant for a premium equal to the purchase price.

The contracts were completed on 2 September 2010. On that date, the first appellants transferred the freehold in the property to Vale and Vale granted a 999-year lease of the property back to the first appellants.

On 7 September 2010, ELS sent a letter to HMRC headed “Acquisition of Land at 14a Tangier Road, Guildford, Surrey, GU1 2DE” summarizing their analysis as to why no SDLT liability arose to the taxpayer on the transactions, in particular, due to the application of section 45(3) and section 71A FA 2003 (and the fact that section 75A did not apply).

On 7 September 2010, ELS also filed three land transaction returns:

On 8 June 2011, and so within the time limit for opening an enquiry, HMRC wrote to the first appellants in the following terms:

Also on 8 June 2011, HMRC wrote to ELS, as the first appellants’ agent, in the following terms:

The letter from HMRC to the first appellants enclosed a copy of the letter to ELS. The letter from HMRC to ELS enclosed a copy of the letter to the first appellants

Also on 8 June 2011, HMRC sent a letter to Vale notifying Vale of their intention to open enquiries into Return B.

There was no further correspondence until 23 October 2013. On that date, following the decision of the First-tier Tribunal in Project Blue, HMRC sent a letter to the first appellants inviting the first appellants to withdraw from the scheme, or, in the alternative, requesting that the first appellants provide information listed in a schedule to the letter.

The letter was in the following terms:

The admissibility of this letter in evidence is disputed by the appellants. They say that the letter is subject to without prejudice privilege. I address this issue later in this decision notice.

On 12 December 2013, ELS wrote to HMRC on behalf of a number of clients, including the first appellants. The schedule to that letter included a list of land transaction returns including Return A. In that letter, ELS argued that the enquiry notices were invalid because no questions were asked and no requests for information were made in those notices. The letter included the following passage:

ELS responded to HMRC’s letter of 23 October 2013 on 9 January 2014 in a letter marked “without the prejudice save as to costs” (Footnote: ²). The letter was in the following terms:

The admissibility of this letter in evidence is also disputed by the appellants. They say that this letter is subject to without prejudice privilege. Once again, I will address this issue later in this decision notice.

HMRC wrote to ELS on 13 January 2014 noting that the first appellants wished to make a payment on account on a without prejudice basis. In that letter, HMRC stated:

On 10 December 2015, ELS wrote to HMRC once again arguing that no valid enquiries had been made by HMRC within the enquiry period. The letter stated:

On 21 April 2016, HMRC responded to ELS’s letter expressing the view that “the enquiry period into those enquiries where closure notices have not been issued remains open”.

On 29 June 2016, following the decision of the Court of Appeal in Project Blue, ELS wrote to HMRC inviting HMRC to close any enquiry. ELS stated:

On 2 August 2016, HMRC responded to ELS noting that HMRC did not object to issuing a closure notice where all enquiries had been completed but that no documents had so far been provided in relation to the enquiries and requesting documents and information.

In response to that letter, ELS wrote to HMRC, on 25 November 2016. In that letter ELS reiterated its position that, following the decision of the Court of Appeal in Project Blue, the first appellants were not liable to pay SDLT. Notwithstanding that assertion, ELS provided various documents whilst stating that they did not consider that HMRC were entitled to the documents as they were “outside of [the] prescribed time period for issuing an assessment”.

On 30 November 2016, HMRC wrote to ELS noting that not all the documentation and information that had been requested had been provided and stating that HMRC could not consider whether to issue a closure notice until that information and documentation had been provided.

On 22 March 2017, ELS responded to the outstanding requests for information. The letter once again raised the argument that, in order for a valid enquiry to be commenced, HMRC were required to make actual requests for documentation and information within the relevant time limit (rather than merely sending a notice of their intention to commence an enquiry). The letter also requested that HMRC issue a closure notice:

“We note that you should have raised any requests for documents and information within 9 months of completion of the acquisition in accordance with legislation and your own guidance. This information is given on a without prejudice basis to our client's contention that you are out of time to raise enquiries.

As this has now been ongoing for a number of years and the Court of Appeal has determined that our client is not the taxpayer for the purposes of this transaction we kindly ask you to issue a closure notice.

On 21 April 2017, HMRC issued two closure notices to the first appellants:

Also on 21 April 2017, HMRC also issued to the first appellants an assessment under paragraph 28 of Schedule 10 in respect of the first appellants’ notional acquisition of the property under section 75A. The assessment confirmed that the notional land transaction was a notifiable transaction for which HMRC had not received an SDLT return and therefore the assessment was subject to the extended time limit under paragraph 31(2A)(b) Schedule 10 FA 2003. The assessment was for SDLT in the amount of £26,240 (before any late payment interest).

HMRC also issued a closure notice to Vale claiming SDLT of £26,240 (before any late payment interest).

On 3 May 2017, ELS wrote to HMRC to appeal against the decisions in the closure notices and the assessment on the grounds that:

On 29 June 2017. HMRC wrote to the first appellants stating their view of the matter and offering a statutory review. This offer was accepted by ELS on 27 July 2017. In a letter dated 8 September 2017, ELS provided further representations.

On 1 November 2017, HMRC issued their review conclusion upholding the assessment and closure notices.

On 30 November 2017, the first appellants notified their appeal to the tribunal. The grounds of appeal were, in summary:

Vale withdrew its appeals against the closure notices issued to it in relation to its participation of the scheme in connection with the acquisitions by the first appellants under the scheme on 25 January 2018.

On 7 March 2018, the tribunal directed that the appeals be case managed and heard together. The appeals were stayed pending the outcome of Project Blue in the Supreme Court. The Supreme Court released its decision on 30 June 2018.

On 24 August 2018, HMRC issued their Statement of Case. In their Statement of Case, HMRC pleaded, inter alia, that:

On 3 March 2020, HMRC applied for the present appeals to be lead cases. On 13 April 2022, the tribunal directed that the appellants should either amend their grounds of appeal or notify their withdrawal of their appeals.

On 24 May 2022, the appellants filed Amended Grounds of Appeal replacing their previous grounds with two grounds of appeal being, in summary:

On 4 July 2022, HMRC amended their Statement of Case to address the new grounds of appeal. The Amended Statement of Case noted (at paragraph 37) that the appellants had not challenged the discovery assessments.

On 31 August 2022, HMRC made a request for further and better particulars in relation to Ground 1 (that HMRC had failed to give notice within paragraph 12 Schedule 10 FA 2003 within the time limit). In that letter, HMRC requested further particulars of the challenges that were made to the validity of the notices to enquire into the returns. HMRC also asked whether the appellants contended that the discovery assessments were procedurally invalid for any reason.

On 16 September 2022, the appellants’ solicitors responded confirming that the appellants were still advancing Ground 1. In relation to that ground, the appellants’ solicitors provided further details as follows:

The appellants’ solicitors wrote to HMRC further on 13 November 2022. In that letter the appellants’ solicitors:

Following their receipt of that letter, HMRC applied on 23 December 2022 to amend their Statement of Case to include an argument that the appellants were estopped from arguing that there were no valid enquiries into the appellants’ returns.

HMRC confirmed, by email dated 1 February 2023, that they were taking the position that paragraph 31A(2) Schedule 10 FA 2003 could apply to permit a 20-year time limit for the issue of the discovery assessments on the grounds that the appellants had failed to file an SDLT return in relation to the notional transaction for the purposes of section 75A FA 2003.

At the hearing, the appellants have not advanced Ground 2 in its Amended Grounds of Appeal – that HMRC could not assess the appellants under section 75A FA 2003 because Vale had withdrawn its appeal and so accepted liability. I have treated this ground as withdrawn.

The appellants advanced Ground 1 of their grounds of appeal – that no valid notice of enquiry had been given to the appellants. HMRC took issue with this ground on the facts. HMRC also raised the alternative argument that the appellants were, in any event, estopped from arguing that no valid enquiry had been opened given the course of conduct and correspondence over several years. In that context, there was an issue between the parties as to whether some of the correspondence between them, on which HMRC relied, was inadmissible as it was subject to without prejudice privilege.

The appellants also sought to challenge the discovery assessments under paragraph 28 Schedule 10 FA 2003 on the grounds that no discovery had been pleaded by HMRC and that any assessment under paragraph 28 was out of time. Once again, HMRC took issue with these arguments, but also argued that the appellants should not be permitted to challenge the discovery assessments as the issue had not been raised until the exchange of skeleton arguments.

The first issue between the parties relates to the validity of any notice of enquiry. In summary, the question is whether the correspondence between the parties was sufficient to constitute notice of an intention to enquire into the both returns (Return A and Return C) within paragraph 12 Schedule 10 FA 2003. If not, then, subject to the issues surrounding whether or not the circumstances give rise to an estoppel to which I will return later in the decision notice, the closure notices issued by HMRC are invalid (Raftopolou v HMRC [2018] EWCA Civ 818 at [51]).

I will begin by setting out the relevant legislation.

Section 76(1) FA 2003 provides for the delivery of SDLT returns in relation to notifiable transactions. So far as relevant, section 76 is in the following form:

The reference to the “purchaser” in section 76 is to a person acquiring a chargeable interest (section 43(4) FA 2003).

The meaning of “notifiable transaction” for the purposes of section 76 FA 2003 is given by section 77 FA 2003. At the relevant time, it provided:

Pursuant to paragraph 12 Schedule 10 FA 2003, HMRC may enquire into an SDLT return by giving notice of their intention to do so to the purchaser within nine months of the filing date (in cases where the return was delivered on or before that date).

At the relevant time, paragraph 12 provided as follows (so far as relevant):

In the case of the transactions in all of these appeals, the filing date was the last day for the filing of a return under section 76(1) FA 2003 i.e. the date 30 days after the effective date of the transaction: paragraph 2 Schedule 10 FA 2003.

Paragraph 13 of Schedule 10 sets out the scope of any enquiry initiated under paragraph 12. It provides:

An enquiry is completed by the issue of a closure notice under paragraph 23 Schedule 10 FA 2003. It provides:

Mr Elliott makes the following submissions on behalf of HMRC.

In summary, Mr Chacko, for the appellants, makes the following submissions.

There was no material dispute between the parties as to the basic principles that should be applied in addressing this question. Both agreed that there is no particular form of notice prescribed by the legislation; all that was required by paragraph 12 Schedule 10 FA 2003 was that the communication was sufficient to inform the first appellants of HMRC’s intention to enquire into the relevant returns (R (Sword Services Ltd) v HMRC [2016] EWHC 1473 per Cranston J at [71])).

The parties also agree that the relevant test is an objective one, namely whether a reasonable taxpayer, in the circumstances of the taxpayer in question, would have understood that HMRC intended to open an enquiry into the particular return. As the Upper Tribunal in Mabbutt describes (Mabbutt [45]):

As can be seen from that extract, Mabbutt concerned a disputed notice to enquire into a return for income tax and capital gains tax under section 8 of the Taxes Management Act 1970 (“TMA”). However, there is no material difference between the provisions of section 9A TMA, which provides for HMRC to give notice of their intention to enquire into such a return, and the provisions of paragraph 12 Schedule 10 FA 2003 which apply for the purposes of SDLT.

That objective test must be applied by reference to the context in which the taxpayer received the disputed notice.

For example, in Mannai, the House Lords found that an error in a notice to exercise a break clause under a lease, which referred to 12 January when the applicable break day was 13 January, did not invalidate the notice. Lord Steyn said this (at Mannai, p772H):

The context may also cause an otherwise valid notice to be treated as invalid. In Barclays, a case concerning notices under the Landlord and Tenant Act 1954, a tenant received an otherwise valid notice agreeing to a new tenancy at the same time as an invalid notice opposing the grant of a new tenancy. The Court of Appeal found that the otherwise valid notice was invalid. This was because the invalid notice opposing the new tenancy was part of the relevant factual context. A reasonable recipient in receipt of the two notices would have been left in reasonable doubt as to whether or not the landlord intended to oppose the new tenancy (Barclays [28]-[34]).

Also – and in a tax context – in Bristol & West, HMRC was not bound by a closure notice which was issued in error where HMRC, having realized its mistake, had informed the taxpayer in advance that it would receive a notice which had been issued in error. The earlier email was part of the relevant factual context in which the disputed closure notice was received (Bristol & West [30]).

Before I turn to the application of these principles to the facts of the present case, I should deal with an issue raised by Mr Elliott in argument. Mr Elliott suggested that I should exercise some caution in applying some of the principles in the cases to which I have referred, in particular, those cases which concern notices given in contexts other than the issue of a notice under paragraph 12 Schedule 10 FA 2003. I do not accept that submission. It seems to me that the principles derived from those cases are of equal relevance in the present context. All the cases concern whether a notice should be regarded as ineffective where the reasonable recipient of the relevant communication would have some doubt about its effect. In any event, the non-tax cases (principally Mannai and Barclays) are regarded as good authority by the Upper Tribunal in GDF Suez and Mabbutt and by the Court of Appeal in Bristol & West. I can see no good reason to diverge from that approach.

That having been said, in my view, the letter of 8 June 2011 sent by HMRC to the first appellants gave sufficient notice of HMRC’s intention to enquire into both returns i.e. Return A and Return C. My reasons are set out in the following paragraphs.

As a starting point, the first appellants received the letter of 8 June 2011 in which HMRC expressed their intention to enquire into a “return” in the singular. That letter referred to and enclosed a copy of the letter to their agents, which referred to an intention to enquire into the “returns” in the plural.

There is a question as to whether the enclosed letter to the agents forms part of the notice to the first appellants. I note that in Mabbutt, the Upper Tribunal took the view that, in very similar circumstances, a letter to the taxpayer’s agent which was enclosed with the letter to the taxpayer, formed part of the same notice. This was on the grounds that any reasonable recipient of the letter to the taxpayer would have understood that the two letters had to be read together (Mabbutt [47]). I take this same approach in this case.

In any event, even if it did not form part of the notice to the first appellants, the letter to the agent in this case must form part of the relevant factual context for the purposes of determining whether or not a reasonable recipient of the disputed notice would have understood that HMRC intended to open an enquiry into both returns.

In argument, Mr Chacko made much of the inconsistencies between the two letters. He submitted, in reliance on cases such as Barclays, that the inconsistences between them could not be resolved. In his submission, the effect of the Upper Tribunal’s decision in GDF Suez was that there must be “no arguable ambiguity” (GDF Suez [117]) and so, if the inconsistencies could not be resolved, the Tribunal must treat the notice to the first appellants as invalid.

In my view, that approach is too narrow. The case law is clear that the relevant test is whether a reasonable recipient, in the circumstances of the taxpayer in question, would have understood that HMRC intended to open an enquiry into a particular tax return (Mabbutt [45]).

Although the communication with the first appellants, which included the copy of HMRC’s letter to the agents, was on its face inconsistent (in referring in the letter to the first appellants to “this return” and in the letter to the agents to “their… returns”), that communication must, as the cases demonstrate, be read in its context. To my mind, that context must include the following facts of which the reasonable recipient must be taken to have been aware:

Against that background, a reasonable taxpayer in receipt of the communication from HMRC would, it seems to me, assume that HMRC intended to enquire into arrangements for the acquisition of the property as a whole and its SDLT treatment. The acquisition of the property was made under a scheme involving pre-planned steps, of which the reasonable recipient would have been aware, and involving the submission of two returns. This is not a case like Barclays where the tenant received two separate communications that were entirely inconsistent. On receipt of the two notices, the reasonable recipient would assume that the letter to the agent was correct in referring to “returns” in the plural and that any error was in the letter to the taxpayer in referring to a return in the singular. For those reasons, a reasonable taxpayer would have been informed of the intention of HMRC to enquire into both returns.

As I have mentioned above, the test is an objective one. The first appellants’ and their agent’s understanding of the letters from HMRC is not strictly relevant. However, I am fortified in my conclusion by the fact that the first appellants and their agents clearly understood the letters in the same way. Although they disputed whether or not the letters constituted a valid notice for other reasons, which have not been pursued, the first appellants and their agents clearly treated the letters as a notice of an intention to enquire into both Return A and Return C for many years. It was only when prompted to amend their grounds of appeal following the decision of the Supreme Court in Project Blue, that the first appellants indicated that they intended to challenge the validity of the disputed notice on this ground (in their Amended Grounds of Appeal filed on 24 May 2022).

The parties argued their cases on the validity of the enquiry notices primarily by reference to the case law authorities. However, I was also referred by the parties to the potential application of section 83(2) FA 2003.

Section 83(2) FA 2003 is part of section 83 which deals with formal requirements of documents relevant to the assessment and collection of SDLT. It provides for certain circumstances in which mistakes and errors in a notice will not be regarded as rendering a notice ineffective. Section 83 is in the following form.

Mr Elliott submits that, even if the notice given to the first appellants was not sufficient to inform the first appellants of HMRC’s intention to enquire into both returns, the reference to a “return” in the singular in the letter to the first appellants was clearly a mistake. However, the letter was substantially in conformity with the requirements of FA 2003 and the intended effect of the letter was reasonably ascertainable by the first appellants. Section 83(2) FA 2003 could therefore apply to correct any error in the notice.

Mr Chacko submitted that HMRC’s argument that section 83(2) FA 2003 might apply to rectify any error in the purported notice was misplaced. In his view, section 83(2) added nothing on the facts of this case. The hypothetical reasonable recipient postulated by section 83(2) would be in no different a position to the reasonable recipient under the case law test as set out in Mabbutt.

Mr Chacko also referred to two decisions of the First-tier Tribunal on the application of section 83(2). They were Coolatinney Developments Limited v HMRC [2011] UKFTT 252 (TC) (“Coolatinney”) and Smith Homes 9 Limited v HMRC [2022] UKFTT 000005 (TC) as illustrations of the relationship between section 83(2) and the principles derived from cases such as Mabbutt. Those decisions are not binding upon me. I have decided this issue by reference to the case law test. Given the parties’ positions – HMRC’s case, that section 83(2) is potentially a broader test than that in Mabbutt, and, on the appellants’ case, that there is no practical difference between them on these facts – I do not need to determine whether there is, in fact, any material difference in the application of the two tests. I do not do so.

I should, however, note in passing one point. This is that the FTT, in its decision in Coolatinney, expressed the view that where the statute itself (through section 83(2)) provided a mechanism for determining whether a notice was ineffective by reason of a mistake in it, recourse could only be had to that statutory provision to resolve the issue (Coolatinney [21]). Neither Mr Elliott for HMRC, nor Mr Chacko for the first appellants, advanced their case on the basis that section 83(2) provided an exclusive mechanism for rectifying errors in a notice under paragraph 12 Schedule 10 FA 2003. I have adopted their approach.

For the reasons that I have given, the letter of 8 June 2011 to the first appellants together with the enclosed copy of the letter to the agents provided sufficient notice to the first appellants of HMRC’s intention to enquire into both Return A and Return C for the purposes of paragraph 12 Schedule 10 FA 2003.

That conclusion decides these appeals in favour of HMRC. However, I have heard full argument on the remaining issues and, in case these appeals progress further, I will express my views on them.

If I had concluded that the notice given to the appellants was not sufficient to inform the first appellants of HMRC’s intention to enquire into both returns, HMRC’s case is that the first appellants are estopped by convention from arguing that the notice failed to identify the returns into which HMRC intended to enquire.

There is, however, a point that I need to address before I turn to the question of estoppel. As I have indicated, the appellants assert that HMRC is not entitled to rely on some of the correspondence in support of this argument because it is subject to without prejudice privilege. The relevant pieces of correspondence are:

Mr Chacko, for the appellants, says that this exchange is subject to without prejudice privilege. Without prejudice privilege is broad. It applies to exclude from evidence all negotiations, whether oral or in writing, genuinely aimed at settlement. In this case, ELS’s letter of 9 January 2014 was in response to the letter from HMRC of 23 October 2013, which was headed “Invitation to Settlement”. Any response to such an invitation must be privileged (R(oao Wildbur) v Ministry of Defence [2016] EWHC 821 Admin per Cranston J [14]-[15], Cutts v Head [1984] Ch 290 per Oliver LJ at page 296). It is not open to HMRC to rely upon non-privileged information (in particular, any reference in the ELS letter to an enquiry being in progress) contained in correspondence to which privilege otherwise applies (Oceanbulk Shipping & Trading SA v TMT Asia Ltd [2010] UKSC 44 per Lord Clarke [25]-[27]).

Mr Elliott, for HMRC, says that none of this correspondence is privileged. A document is only privileged if it is genuinely part of negotiations towards a settlement. It does not apply to a communication which is simply an assertion of one party’s case or a criticism of the other party’s case (Galliford Try Construction Limited v Mott Macdonald Ltd [2008] EWHC 603 TCC (“Galliford Try”) [5.2(c)]). In this case, neither letter was attempting to negotiate the dispute between the parties. The fact that the ELS letter was headed “without prejudice” makes no difference; it was not part of communications aimed at settling the dispute (Galliford Try [5.3(a)]).

Notwithstanding the breadth of the principle, on this issue, I agree with Mr Elliott.

The cases to which I have been referred indicate that the without prejudice rule is based, at least in part, on public policy arguments in favour of promoting settlement and compromise and so should apply to exclude all negotiations genuinely aimed at settlement whether oral or in writing from being given in evidence. That policy would not extend to these communications. They were not made with any intention of opening or furthering negotiations or seeking any form of compromise of the dispute. HMRC’s letter is simply a statement of its own position, inviting the appellants to capitulate (Galliford Try [5.2(c)], referring to Buckinghamshire County Council v Moran [1990] Ch 623). ELS’s letter of 9 January 2014 contains no indication of any intention to negotiate or compromise at all.

I now turn to the question of estoppel by convention.

The parties made their submissions on this issue by reference to the decision of the Supreme Court in Tinkler v HMRC [2021] UKSC 39 (“Tinkler”). I will begin by summarizing the facts of, and the principles that I draw, from that case.

Tinkler concerned a notice of intention to enquire into a return for the purposes of income tax and capital gains tax under section 9A TMA. A notice to of intention to enquire into such a return may be served by delivery to the taxpayer at the taxpayer’s “last known place of residence or his place of business or employment” (section 115 TMA).

In Tinkler, the relevant notice was sent to Mr Tinkler’s previous address even though HMRC were aware of the new address. A copy of the notice was also sent to Mr Tinkler’s accountants. The accountants’ response included a statement “as the return is now the subject of a section 9A TMA enquiry”. The enquiry continued on the basis that it had been validly opened. At the conclusion of the enquiry, HMRC issued a closure notice in which they denied Mr Tinkler the use of a loss that he had claimed, and sought to impose additional tax on Mr Tinkler of approximately £700,000.

Mr Tinkler appealed. In addition to his arguments on the substantive issues, Mr Tinkler amended his notice of appeal to the FTT to include an argument that the notice under section 9A was invalid (and so the closure notice was invalid) because it had not been properly served on him.

The only issue before the Supreme Court in Tinkler was whether Mr Tinkler was estopped by convention from denying that HMRC had validly opened an enquiry. The Supreme Court decided that he was. In his judgment, with which all the members of the Supreme Court agreed, Lord Burrows approved the principles which govern the application of estoppel by convention as set out by Briggs J in HMRC v Benchdollar Ltd [2019] EWHC 1310 (Ch) subject to a refinement derived from the decision of the Court of Appeal in Dixon v Blindley Heath Investments Ltd [2015] EWCA Civ 1023 (Tinkler [42]-[50]).

Those principles are conveniently set out in the decision of Master Clark in Brierley v Otuo [2022] EWHC 1530 (Ch) at [38]:

Applying these principles to the facts in Tinkler, Lord Burrows found:

Mr Elliott, for HMRC, says that the present case is on all fours with Tinkler:

Mr Chacko raises various challenges to that analysis. I will deal with them in turn.

Mr Chacko submits that there was no common assumption that could found an estoppel in this case. As a starting point, the assumption on which HMRC rely – that a valid enquiry had been opened – is too vague. He relies on the decisions in Republic of India v Indian Steamship Limited [1997] UKHL 40 (“The Indian Endurance”)and HIH Casualty & General Insurance Limited v Axa Corporate Solutions [2002] EWCA Civ 1253 in support of a submission that, in order to found an estoppel, that a common assumption must be specific to the issue at hand. So, for example, in The Indian Endurance, a shared common assumption that a smaller claim was preceding outside the UK (at Cochin) and that the larger claim would be brought elsewhere was not sufficient to bar a claim before the UK courts. The defendant would have had to made clear that “in future proceedings no plea or defence on the basis of a judgment in Cochin would be raised whatever the outcome of the proceedings in Cochin”.

Mr Elliott points out that, at various points in his judgment in Tinkler, Lord Burrows refers to the relevant assumption as being that “a valid enquiry had been opened”. This is the same common assumption for which HMRC argue in this case. In response, Mr Chacko says that the common assumption in Tinkler was, in effect that notice had been received by Mr Tinkler at the correct address. I do not accept that submission. Lord Burrows refers to the common assumption on several occasions in his judgment in exactly the same terms. See, for example, at [63] and perhaps most importantly in his conclusion at [84]. Furthermore, Lord Briggs in his short concurring judgment uses exactly the same phrase to describe the common assumption (Tinkler [91]). For these reasons, I conclude that, at least in principle, an assumption that a valid enquiry had been opened is sufficiently specific to found an estoppel.

Mr Chacko also says that it cannot be said on the facts of this case that the first and second requirements for an estoppel by convention as set out by Lord Burrows in Tinkler are not met. At no point did the first appellants “cross the line” by making it clear to HMRC that they assented to a common assumption that a valid enquiry had been opened. Nor could it be said that the first appellants assumed any responsibility for the common assumption.

On this issue, I agree with Mr Chacko.

As I have described, it is clear from the decision of Lord Burrows in Tinkler that, in order to found an estoppel by convention, it is not enough that the common assumption is identified by both parties. There must be a “crossing of the line” sufficient to show that the person against whom the estoppel is raised assented to the common assumption. Also, the person against whom an estoppel is raised must have assumed some element of responsibility for the common assumption, in the sense of conveying to the other party an understanding that the person against whom an estoppel is raised expected the other party to rely upon it.

Lord Burrows says this (at Tinkler [52]):

The point is reiterated by Lord Briggs in his short judgment (Tinkler [88]-[89]).

Whilst it is true that the first appellants or their agents refer to “enquiries” in their letters of 12 December 2013, 9 January 2014, 10 December 2015, 29 June 2016 and 22 March 2017, I cannot read any of this correspondence as implying an acceptance, on the part of the first appellants, that HMRC had opened a valid enquiry.

There is no unqualified statement in this case by the first appellants or their advisers, such as that in Tinkler, where the taxpayer’s agent stated in correspondence with HMRC its assumption that a valid enquiry had in fact been opened. Indeed, throughout this period, the first appellants continued to assert that no valid enquiry had been opened. This was on the grounds that were articulated in the letter from ELS dated 12 December 2013. This argument was still being made in the letter of 22 March 2017, only shortly before the issue of the closure notices in April 2017. The first appellants no longer rely on those grounds. However, against that background, the reference to the continuing enquiries in this correspondence can only be read as referring to enquiries which HMRC regarded as validly opened and on-going. I cannot read into them an acceptance on the part of the appellants that the enquiries had otherwise been validly opened.

Furthermore, there is no point in the correspondence or the conduct of the enquiry that the appellants could be said to have “assumed responsibility” for the assumption in terms of conveying to HMRC that the appellants intended and expected HMRC to rely upon it. Nor did the appellants or their agents – as the taxpayer’s agents did in Tinkler – proceed to rely upon that assumption in future dealings with HMRC. Whilst the appellants provided some documentation to HMRC, this was on the basis as expressed in the letter from ELS of 25 November 2016 that there was no liability and that HMRC was out of time to issue an assessment (because no valid enquiry had been made).

Mr Elliott submits that notwithstanding these objections to the validity of any enquiry, there remained a common assumption, on which HMRC relied, that a notice of intention to enquire had been given in relation to both returns and that the parties proceeded on that basis. I assume by this argument he means that the common assumption was not that a valid enquiry had been opened but rather that some form of notice – whether or not it constituted a valid enquiry notice – had been received by the appellants in relation to both returns. As will be apparent from my comments in relation to the question of whether a valid notice was in fact received, I agree that was the case. But if this is the assumption to which Mr Elliott refers, there is no statement on the part of the appellants or their representatives by which they “crossed the line” and indicated to HMRC that they assented to that more specific assumption. Even if some of their actions in pursuance of the enquiry could be regarded as “crossing the line”– which I do not agree - the appellants did not assume responsibility for the common assumption. I do not accept, in a case such as this, where the appellants are actively disputing the validity of an enquiry – albeit on a basis that is subsequently withdrawn – that by continuing to take action in pursuance of the enquiry to protect their positions, they should be regarded as representing to HMRC in a manner sufficient to ground an estoppel, which can be relied upon by HMRC, that there is no other basis on which the appellants might challenge the validity of the notice.

For these reasons, I agree with Mr Chacko that the requirements for an estoppel by convention adopted by the Supreme Court in Tinkler are not met in this case. The first appellants did not “cross the line” in relation to a common assumption or assume responsibility for a common assumption by affirming the assumption to HMRC in a manner that they intended and expected HMRC to rely upon it. Accordingly, if I had found that HMRC had not validly issued a notice to enquire into both returns, I would have concluded that the correspondence and conduct of the appellants was not sufficient to support a claim of estoppel by convention by HMRC.

Mr Chacko also raised a technical challenge to the application of an estoppel in this case.

Mr Chacko says that an estoppel by convention cannot apply in these circumstances. The point at issue in this case is whether a valid notice to enquire into a return has been issued. The question goes to the jurisdiction of the tribunal. If there is no valid notice, the closure notice is not valid and the tribunal does not have jurisdiction to hear the appeal (Raftopoulou [51]). Just as it is not possible for the parties to agree that a statutory tribunal has jurisdiction when it does not, it is not possible for estoppel to be used to expand the jurisdiction of a statutory tribunal when it would not otherwise have such jurisdiction. Mr Chacko relies on the decisions of the House of Lords in Essex County Council v Essex Incorporated Congregational Church Union [1963] AC 808 per Lord Reid at pages 820-821 and J&F Stone Lighting and Radio v Levitt [1947] AC 209 at page 216 in support of this submission.

The point is similar to the principle that estoppel cannot be used to undermine a statutory protection. This is the principle underlying the decision of the Court of Appeal in Keen v Holland [1984] 1 WLR 251. That case concerned the operation of the Agricultural Holdings Act 1948. In response to an argument that it was possible for the statutory protections afforded by the Act to be ousted by the operation of estoppel, Oliver LJ says this at page 261C:

There are similar statements that the principle in the judgments of the Court of Appeal in Daejan Properties Ltd v Mahoney [1995] 28 HLR 498 (per Bingham LJ at page 505 and Hoffman LJ at page 511).

In response to this point, Mr Elliott relies on the decision of the Supreme Court in Tinkler. He points out that, in that case, the failure of HMRC to issue a valid notice did not prevent the Supreme Court from concluding that the taxpayer was estopped from denying that a valid enquiry had been opened. Furthermore, a conclusion that the appellants are estopped from denying that valid enquiries were opened does not impermissibly outflank a statutory protection – in particular, as in both Tinkler and this case, where the taxpayer knew of HMRC’s intention to enquire (Tinkler [65]).

This point is directly addressed by Lord Burrows in Tinkler. In Tinkler, Lord Burrows distinguishes Keen v Holland on the grounds that section 115 TMA is permissive. It simply provides a method by which a notice can be given. HMRC and Mr Tinkler could have agreed that notice could be given in another way. It followed that estoppel by convention could apply. Lord Burrows says this at [81]-[82]:

An estoppel could therefore apply in a case, like Tinkler, where it was otherwise open to the parties to agree a means of service of the notice. However, that leaves open the question in a case such as this as to whether estoppel can apply in a case where the taxpayer’s argument is that no notice has been given at all. In such a case, I can see some force in Mr Chacko’s argument that the question becomes simply whether or not a valid notice has been given and there is no room for the operation of estoppel. However, I have decided that no estoppel arose on the facts of this case. I do not need to decide this point and I do not do so.

For the reasons that I have given above, in my view, the requirements for an estoppel by convention are not met in this case.

As I have mentioned above, HMRC issued discovery assessments in relation to each charge to tax under s75A FA 2003 to protect their position if the notices to enquire into the returns should for any reason be invalid.

The appellants seek to challenge the discovery assessments on three grounds.

Before I address these issues, I will first set out the relevant legislation.

In certain circumstances, HMRC has power to make a discovery assessment outside the normal enquiry process. The relevant provision is in paragraph 28 Schedule 10 FA 2003. It provides:

There are restrictions on HMRC's power to make a discovery assessment in cases where the purchaser has made a return. These restrictions are contained in paragraph 30 Schedule 10 FA 2003. Paragraph 30 provides:

These provisions relating to SDLT are similar - but not identical - to provisions which allow HMRC to make discovery assessments for income tax and capital gains tax, which are contained in section 29 TMA, and corporation tax, which are found in paragraphs 41 to 45 of Schedule 18 to the Finance Act 1998 (“FA 1998”). Some of the authorities to which I was referred by the parties concern the application of these other provisions. Although the wording of section 29 TMA and paragraphs 28 and 30 Schedule 10 FA 2003 differ in some respects, the parties regarded the principles which are outlined in those cases as applicable equally to both sets of provisions. I have adopted the same approach.

The time limit for the making of an assessment is set out in paragraph 31 Schedule 10 FA 2003 and is usually four years from the effective date of the transaction: paragraph 31(1). There are, however, various exceptions to that general rule. Where the loss of tax is attributable to carelessness on the part of the taxpayer, the time limit is extended to six years from the effective date of the transaction: paragraph 31(2). The time limit is further extended to 20 years from the effective date of the transaction in various circumstances set out in paragraph 31(2A). Those circumstances include: where the loss of tax is brought about deliberately by the taxpayer; and where the loss of tax is attributable to the failure of the taxpayer to comply with its obligations under various provisions including section 76(1) FA 2003 (the obligation to file a land transaction return).

Paragraph 31 is, so far as relevant, in the following form.

As a preliminary point, HMRC challenge the appellants’ right to contest the discovery assessments. I heard argument on this point at the beginning of the hearing, but decided that I would reach a conclusion on the point as part of this decision, and having heard the arguments on the three issues to which I have referred above.

In short, Mr Elliott, for HMRC, says that, although the appellants’ original grounds of appeal dated 30 November 2017 asserted that the discovery assessments were out of time, the amended grounds of appeal dated 24 May 2022 did not contain any challenge to the discovery assessments. HMRC expressly noted in their Amended Statement of Case of 4 July 2022 that no challenge was made to the discovery assessments. When the issue was raised with the appellants, the only issue raised in the appellants’ solicitors’ response related to the conditions in paragraph 30 Schedule 10 FA 2003. This point was not taken further. HMRC considered that the matter had been withdrawn. The first time that HMRC were aware that the appellants had resurrected their challenge to the discovery assessments was at the time of the exchange of skeleton arguments. HMRC say that, given how late this issue has been raised, the appellant should not be entitled to run these arguments before the tribunal.

I do not accept Mr Elliott’s arguments on these issue. As Mr Chacko pointed out, the validity of a discovery assessment is an issue which it is incumbent on HMRC to make a positive case. The burden is clearly on HMRC (Burgess [38]). The appellants are not required to plead a positive case. HMRC is not entitled to assume the absence of any reference to this issue in the grounds of appeal is some form of concession or waiver of the point (Burgess [49]). In any event, HMRC were clearly on notice that they might be put to proof on various issues relating to the discovery assessments at the very latest following their receipt of the appellants’ solicitors’ letter of 13 November 2022. That letter included a reference to both the discovery issue and the time limit issue. Mr Elliott invited me to conclude that the appellants had withdrawn or conceded the point. However, any withdrawal or concession would have to be clear and unambiguous (Burgess [44]), and I cannot infer from the relevant correspondence any form of waiver or concession on the part of the appellants.

I have therefore concluded that the appellants should be permitted to raise these issues before the tribunal.

As regards the question of discovery, Mr Elliott applied to admit the evidence of Mr Hemmings on this issue. He says that HMRC should be entitled to submit the additional evidence given the manner in which this point has arisen, which he described as a “planned ambush” by the appellants.

I do not accept Mr Elliott’s description. However, I have decided to admit Mr Hemmings’s evidence. It has become clear that there has been some confusion over the issues which are to be argued before the Tribunal in this case. It would be fair and just for me to hear all of the arguments and all of the evidence at this time. The evidence is unchallenged.

All that is required for HMRC to prove a “discovery” is for HMRC to show that an officer of the board came to a conclusion that there is a possible insufficiency. This is a subjective test (HMRC v Tooth [2021] UKSC 17, [72] and [78] – [80]). It is Mr Hemmings’s unchallenged evidence that, having been instructed to issue closure notices for the enquiries that are relevant to these appeals, he came to the view that the appellants had failed to submit returns under section 76(1) FA 2003 in respect of the notional transactions under section 75A FA 2003 and that there was a risk of “significant losses” of SDLT as a result. He then arranged for the discovery assessments to be issued.

There is, of course, an issue as to whether Mr Hemmings’s analysis is correct – this is the time limit issue. However, discovery in this case is proved by the evidence of Mr Hemmings.

The time limit issue arises from the fact that the discovery assessment in this case was issued on 21 April 2017. The assessment was made outside the usual four-year time limit for the making of a discovery assessment (paragraph 31(1) Schedule 10) and outside the six-year time limit that applies where the loss of tax is attributable to carelessness on the part of the taxpayer or a related person (paragraph 31(2) Schedule 10). Instead, HMRC says that the discovery assessments are made in time because the 20-year time limit in paragraph 31(2A) Schedule 10 applies.

In summary, Mr Elliott, for HMRC, submits:

Mr Chacko disagrees. He says that it is possible to treat the SDLT return submitted by the first appellants in respect of their purchase of the property from the third party vendor (Return A) as the relevant return in respect of the notional land transaction (so that paragraph 31(2A)(b) does not apply). He asserts that this conclusion follows from the judgment of Lord Hodge in Project Blue.

The relevant passage from the judgment of Lord Hodge in Project Blue is found at [81] to [84] and is as follows:

Mr Chacko also refers to a decision of the Upper Tribunal in Brown v HMRC [2022] UKUT 298 (“Brown”). Brown was a case involving a different scheme to avoid SDLT on the purchase of residential property, which was also based on sub-sale relief under section 45(3). Under that scheme, the purchasers of a property funded an unlimited company by subscribing shares. The company then purchased the property and transferred the property to the purchasers by distribution in specie. The Upper Tribunal decided the case by reference to a point on the construction of section 45(3) FA 2003. However, one issue that was before the Upper Tribunal concerned the scope of the determination made by HMRC under paragraph 25 Schedule 10 FA 2003. On that issue, the Upper Tribunal expressed the view that a determination made in respect of the acquisition of the chargeable interest by the purchaser was capable of meeting the requirements of paragraph 25 whether the charge to tax was under section 45(3) FA 2003 or under section 75A FA 2003.

Mr Chacko refers in particular to a passage at the end of the decision of the Upper Tribunal (at [88] to [89]) in which the Upper Tribunal considers whether its conclusion on this latter point is consistent with the decision in Project Blue. The Upper Tribunal says this:

As I have mentioned above, the Upper Tribunal decided the case by reference to the construction of section 45(3) FA 2003. So these statements are strictly obiter. However, Mr Chacko says that the effect of the decision of the Supreme Court in Project Blue, as interpreted by the Upper Tribunal in Brown, is that a return is made in relation to the relevant land transaction, which is the acquisition of a chargeable interest by the first appellants from the third party vendor. It does not matter whether that acquisition takes place pursuant to an actual transaction for SDLT purposes or pursuant to a notional transaction under section 75A. It follows, in this case, that the return made in respect of the sale from the third party vendor to the first appellants (i.e. Return A) should properly be regarded as a return in relation to the underlying land transaction, whether that is an actual transaction or a notional land transaction under section 75A FA 2003. As a result, the obligation in section 76(1) has been fulfilled and paragraph 31(2A)(b) does not apply and the discovery assessments are out of time.

It is not clear to me that Lord Hodge’s words in Project Blue extend as far as Mr Chacko suggests. In the passage from his judgment (at Project Blue [81]-[84]), Lord Hodge rejects an argument that HMRC was not entitled to amend the return – equivalent to Return A in this case – relating to the purchase of property from the third party vendor in order to impose a liability to SDLT on a notional land transaction under section 75A (because the return related to a separate transaction which had been disregarded under section 45(3) FA 2003). Instead, Lord Hodge takes the view that it was open to HMRC as part of the enquiry into that return to impose a liability on a notional transaction under section 75A. He reaches that view by reference to the scope of the enquiry into that return, which by virtue of paragraph 13 Schedule 10 extends to anything included in the return. The return contained information regarding the underlying transaction (i.e. the sale from the third party vendor to the purchaser). HMRC were entitled to enquire into the consequences of that transaction, which as it was part of a wider scheme, included the possibility of a charge under section 75A.

The passage therefore refers to the scope of HMRC’s enquiry into a return. As Lord Hodge states, at the end of paragraph [83] of his judgment, the scope of HMRC’s powers to enquire into the return relating to the actual transaction is not limited by any obligation on the purchaser to submit a return in relation to the notional land transaction.

This observation does not, however, exclude Mr Chacko’s point. As I understand it, his analysis – consistent with the Upper Tribunal’s comments in Brown (Brown [88]-[89]) – is as follows.

I should note, however, that the transactions in Project Blue (and in Brown) took place under a slightly different regime for the notification of land transactions in SDLT returns. In particular, they took place before the introduction in Finance Act 2008 of a new section 77 FA 2003, which included the specific reference to “a notional land transaction under section 75A” (in paragraph (d) of subsection (1)) in the list of notifiable transactions. The effect of that change, in combination with the obligations to file returns in section 76 FA 2003, was to introduce a specific obligation to file an SDLT return in respect of a notional land transaction created by section 75A FA 2003. There was no such specific reference in the prior legislation.

The key question is whether the changes made by the Finance Act 2008 affect the analysis. I have decided these appeals by reference to the validity of the notices to enquire into the returns. However, if I am wrong on that issue, I would agree with Mr Chacko that the Finance Act 2008 changes do not affect the analysis. Return A was a return in respect of the notional transaction under section 75A. Paragraph 31(2A)(b) Schedule 10 FA 2003 does not apply. The discovery assessments are therefore out of time.

The final point that I need to address is the pleading issue.

In short, the appellants argue that section 75A can only apply where both section 45(3) FA 2003 and section 71A FA 2003 apply to the transactions in the scheme. HMRC has continued to plead that section 45(3) applies. Mr Chacko refers to the Upper Tribunal’s decision in Brown (Brown [53]) as authority for the proposition that section 45(3) does not apply on the basis that the transactions are self-cancelling (applying BMBF)

On this point, I agree with Mr Elliott. HMRC’s case before the tribunal has been that section 75A FA 2003 applies to the transactions. HMRC has not sought to argue that the financing transactions are self-cancelling applying BMBF and so section 45(3) cannot apply. HMRC’s case rests on the similarity of the transactions in this case with those in Project Blue. The analysis follows that of the Supreme Court in Project Blue.

For the reasons that I have given, HMRC gave notice of intention to enquire into both returns (Return A and Return C) under paragraph 12 Schedule 10 Finance Act 2003. The closure notices are therefore valid.

I dismiss these appeals.

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.