Barclays Bank PLC v The Commissioners For HMRC

This is an appeal against a decision of the First-tier Tribunal (Tax Chamber) (“the FTT”) released on 26 April 2024 (“the Decision”). It concerns the tax treatment of loan relationships entered into by the Appellant (“BBPLC”) in 2008 at the height of the financial crisis. In particular, whether the FTT was entitled to find that the accounts of BBPLC were not compliant with generally accepted accounting practice (“GAAP”).

BBPLC issued interest bearing debt instruments called reserve capital instruments (“RCIs”) on 27 November 2008 with a face value of £3bn expressed to be issued at par. On 31 October 2008, its parent company Barclays PLC (“Barclays”) had issued 5-year warrants expressed to be issued for a nominal consideration to subscribe for 1,516,875,236 of its ordinary shares (“the Warrants”) at the average closing price of those shares on 29 and 30 October 2008.

BBPLC received £3bn when the RCIs were issued. It accounted for the transactions on the basis that £800m was attributable to the Warrants issued by Barclays and this was treated as a capital contribution from Barclays to BBPLC. The balance of £2.2bn was treated as the fair value of BBPLC’s liability under the RCIs. The discount of £800m was treated as an accruing discount being written off over the expected lifetime of the RCIs. The accruing discount was recognised as a debit arising in relation to a loan relationship for corporation tax purposes.

The FTT found that this accounting treatment was not in accordance with GAAP. It concluded that £3bn was paid to BBPLC for the RCIs and that £3bn was therefore the fair value of the RCIs which should have been included in its accounts as such. No part of that sum should have been allocated to the Warrants. In the circumstances, the RCIs were not issued at a discount and there was no debit to recognise for corporation tax purposes.

Mr Prosser KC and Mr Henderson for BBPLC submitted that the FTT erred in law in reaching that conclusion. The only reasonable conclusion available to the FTT was that as a matter of substance and economic reality the £3bn was paid for the RCIs and the Warrants. In the alternative, it was submitted that in reaching its conclusion the FTT took into account irrelevant factors and failed to take into account all relevant factors.

Ms Wilson KC and Mr Simpson for HMRC submitted that the FTT was entitled to make the findings it did as to the accounting treatment and that we should not interfere with those findings.

We are grateful to all counsel and the parties’ legal teams for the quality of their written and oral submissions which have assisted us greatly.

The relevant statutory provisions were at all material times contained in Chapter II Part IV Finance Act 1996 (“FA 1996”). This dealt with loan relationships and how profits and losses arising from a company's loan relationships were to be brought into account for corporation tax purposes. There are no issues of construction in relation to the relevant provisions and we can summarise the statutory framework quite briefly.

Chapter II Part IV FA 1996 provides that all profits, gains and losses arising to a company from its loan relationships shall be taken into account for the purposes of corporation tax.

Section 81 FA 1996 defines a “loan relationship”. Essentially, a company has a loan relationship whenever it stands in the position of a creditor or debtor as respects a money debt arising from a transaction for the lending of money. The RCIs therefore amounted to a loan relationship.

Section 84(1)(a) provides that the credits and debits to be brought into account by a company for corporation tax purposes in respect of its loan relationships are the sums which when taken together “fairly represent”, for the relevant period, all profits, gains and losses (disregarding interest, charges and expenses) arising to the company.

Section 85A provides that subject to section 84(1), a company is required to bring into account the amounts that, in accordance with generally accepted accounting practice, are recognised in determining the company’s profit or loss for the period. If a company does not draw up accounts in accordance with GAAP, described as “correct accounts”, then the loan relationship provisions apply as if it had drawn up correct accounts.

The FTT identified the issues in the appeal at [148]. At this stage, we can summarise the issues before the FTT as follows:

In the event, the FTT did not determine Issues (2), (3) and (5) because it found that the £3bn was paid only for the RCIs. There is some doubt as to whether the FTT determined Issue (4), but as we explain below, that issue is not before us. We are only concerned with Issue (1) on this appeal. If we were to allow the appeal in relation to Issue (1) and the decision is re-made in favour of BBPLC, either by us or on remittal to the FTT, then the FTT would also have to determine the remaining issues.

There was a considerable amount of documentary evidence before the FTT, but very little witness evidence as to the circumstances in which the RCIs and the Warrants were issued. There was also a statement of agreed facts. The FTT had expert evidence on accounting practice and valuation. In the event, the FTT did not need to consider much of the expert evidence which was mainly relevant to Issues (2) – (5).

The circumstances in which the RCIs and the Warrants were issued has been the subject of considerable previous litigation. Most relevant for present purposes is the judgment of Waksman J in PCP Capital Partners LLP & Anor v Barclays Bank PLC [2021] EWHC 307 (Comm) (“PCP”).

We summarise the statement of agreed facts which the FTT adopted at [24] to [52] of the Decision in the following paragraphs.

At the height of the financial crisis, the Financial Services Authority (“the FSA”) instructed the major UK banks to increase their Tier 1 capital ratios. Barclays’ response was to raise additional capital from new strategic investors. Those investors were the state of Qatar through Qatar Holding LLC (“Qatar”) and Sheikh Mansour of Abu Dhabi through PCP Gulf Invest 2 Limited and PCP Gulf Invest 3 Limited (“PCP”) (together, “the Subscribers”).

On 31 October 2008 various agreements were entered into which were described by the FTT as the RCI Subscription Agreements and the Warrant Subscription Agreements.

Pursuant to the RCI Subscription Agreements (“the RSAs”), Qatar and PCP separately agreed with BBPLC to each subscribe for £1.5bn of RCIs. The RCIs carried interest at 14% pa to 15 June 2019 and 13.4% plus LIBOR thereafter. There was no fixed redemption date but BBPLC could redeem all but not some of the RCIs from 15 June 2019. The obligation to subscribe was conditional on the approval of Barclays’ shareholders. BBPLC undertook to pay commissions of 2% to the Subscribers, totalling £60m.

Pursuant to the Warrant Subscription Agreements (“the WSAs”), Qatar and PCP separately agreed with Barclays each to acquire for £0.76, 758,437,618 warrants to subscribe for 758,437,618 ordinary shares in Barclays. The right to exercise the Warrants was conditional on Barclays shareholder approvals and BBPLC receiving full payment in respect of the RCIs. The Warrants were exercisable at any time after those conditions were satisfied until 31 October 2013. The exercise price for each Warrant was 197.775p which was the average closing price of Barclays ordinary shares on 29 and 30 October 2008. Barclays was the listed entity. BBPLC was its subsidiary and was not listed.

These agreements were negotiated on arm’s length terms. Barclays was not forced to deal with the Subscribers. It could have accepted government funding.

The capital raising also involved the issue by BBPLC of mandatorily convertible notes (“MCNs”). We understand that £2.8bn of MCNs were issued to the Subscribers and £1.5bn of MCNs were issued to other investors. The total raised in the capital raising was therefore £7.3bn. The treatment of the MCNs is not the subject of this appeal.

On 31 October 2008, Barclays issued the Warrants and received the total consideration of £1.52.

Following discussions between Barclays and its institutional investors, the Subscribers each agreed to make available to those investors £250m of RCIs, without Warrants. The Subscribers remained obliged to subscribe and pay for those RCIs in the event the institutional investors did not fully subscribe for the RCIs.

Shareholder approvals were obtained on 24 November 2008. The RCIs were issued on 27 November 2008 and were listed on the Official List of the London Stock Exchange. BBPLC received £1.25bn from each of Qatar and PCP and £500m from the institutional investors.

BBPLC was beneficially entitled to the proceeds of the RCIs. No part of it was held on behalf of Barclays and no part was paid to Barclays.

In the event, all the Warrants were exercised by February 2013 and the RCIs were redeemed in full for £3bn on 15 June 2019.

The statement of agreed facts also dealt with accounting and valuation aspects of these transactions. For the purposes of the group and solus accounts of Barclays and BBPLC, the proceeds were apportioned between the RCIs and the Warrants on the basis of what was estimated to be fair values at a commitment date of 31 October 2008. £800 million was attributed to the Warrants and the balance of £2.2 billion was attributed to the RCIs. We were told that this was on the basis that the fair value of the Warrants was £800m, but there was no liquid market for the RCIs on that date. The fair value of the RCIs was therefore a balancing figure. This was described in the evidence to the FTT as a “warrants first approach” and was the subject of Issue (3) which was not determined by the FTT.

BBPLC accounted for cash received of £3bn as representing £2.2bn for the RCIs and a capital contribution from Barclays of £800m. It recognised the annual interest on the RCIs as an expense in its income statement. It also charged the discount of £800m over the lifetime of the RCIs which it expected to redeem on 15 June 2019. Barclays accounted for an investment of £800m in BBPLC represented by equity capital.

BBPLC brought an accrued discount of £7,253,036 into account as a loss arising from its loan relationships in its accounts for the first relevant accounting period ending 31 December 2008. It did not bring into account any part of the £800m as a profit.

The FTT went on to make further findings of fact at [53] – [144] of the Decision. We summarise the FTT’s further findings of fact in the following paragraphs.

The FTT first noted certain matters from the judgment of Waksman J in PCP. PCP had sued Barclays on the basis of representations made by Mr Roger Jenkins of Barclays to the effect that it would obtain the same deal as Qatar in the capital raising. In fact, a further payment of £280m was to be made to Qatar following the above transactions pursuant to an advisory services agreement (“ASA 2”). Barclays had also provided an unsecured loan of $3bn to Qatar as part of its overall deal with Barclays. As to the relevance of the judgment of Waksman J in PCP, the FTT noted at [16] that it was not bound by those findings but treated his findings as a starting point in its consideration of the evidence. No criticism is made of the FTT for taking this general approach but there is some criticism by BBPLC of the reliance placed by the FTT on certain aspects of the judgment.

At the time of the financial crisis, Barclays was determined not to be subject to a UK government bailout. There had been an earlier capital raising of £4.5bn by Barclays in June 2008 in which Qatar invested £2bn (“CR1”). We understand that this was by way of a subscription for share capital. At that time, Qatar and Barclays had also entered into an advisory services agreement whereby Qatar agreed to provide advisory services to Barclays in return for advisory fees.

Negotiations took place with the Subscribers in October 2008. On 8 October 2008, HM Treasury announced a bank recapitalisation scheme. Mr Jenkins of Barclays spoke to Qatar about a further capital raising (“CR2”) potentially involving warrants.

By 11 October 2008, the issue of preference shares with detachable equity warrants was being considered. There was a view within Barclays that they would be unable to raise £3-4bn without accompanying warrants.

Barclays publicly announced on 13 October 2008 that it intended to raise £6.5bn additional Tier 1 capital without calling on government funding.

By 14 October 2008, the Barclays finance committee considered that RCIs would be more attractive to investors than preference shares. The option of attaching warrants to the RCIs was being investigated. RCIs without warrants would be expected to pay a coupon of 13-14% whereas those with a warrant would be expected to pay approximately 10%.

By 17 October 2008, Qatar had agreed in principle to purchase £1.5bn of a £3bn RCI issue, together with £500m equity at a later stage. However, on 20 October 2008 a structure involving RCIs carrying a 10% coupon with detachable warrants, or a higher coupon with no warrants was being discussed with PCP. Board minutes show that Barclays preferred the latter “vanilla” structure.

By 21 October 2008, Barclays was moving towards an RCI deal with Qatar with no warrants. Warrants were described as “very precious” and would be “guarded jealously”. However, subsequent presentations referred to Qatar and PCP receiving £1.5bn of warrants and it appears that Barclays considered that Qatar was asking too much. It wanted a coupon of 14% on the RCIs and the warrants. PCP meanwhile wanted £3bn worth of warrants.

Term sheets were sent out on 24 October 2008, for RCIs and warrants together with the MCNs. There were discussions with Qatar about further fees being paid by Barclays to Qatar and about a $3bn loan, with Qatar threatening to walk away from the deal. For a while, the Libyan Investment Authority was a potential subscriber, but had fallen out of the picture by 31 October 2008.

A Barclays board meeting on 26 October 2008 described Qatar and PCP as wanting to invest £6bn in total. £3bn of this was by way of RCIs with warrants attached.

On 30 October 2008, Qatar raised the need to get a “blended” share price of 130p per share following falls in the share price since CR1 in June 2008. It seems that by this it meant a deal which gave an overall share price of 130p combining its investment in CR1 and CR2. On the same day, a $3bn unsecured loan by Barclays to Qatar was agreed.

The deal was announced on 31 October 2008. The RSAs and WSAs were dated the same date. We were taken to Barclays’ public announcement of the capital raising which referred to the issue of £3bn RCIs and £4.3bn MCNs. It was said that “in conjunction” with the issue of RCIs, Qatar and PCP had subscribed for the Warrants. The announcement stated that Barclays had previously announced a plan to raise £6.5bn of Tier 1 capital of which £3bn would be in the form of preference shares and the balance in ordinary shares. It now stated that the objective of the plan to raise Tier 1 capital by the issue of preference shares would be met by issuing the RCIs and that MCNs would be issued rather than ordinary shares.

By 10 November 2008, Barclays’ institutional investors were expressing some discontent with the proposed deal, which led to £500m of the RCIs being allocated to those investors. This was achieved by the Subscribers signing “concession letters” which we describe below.

The FTT noted the findings of Waksman J in PCP that Barclays was desperate and would go to extraordinary and unusual lengths to obtain the necessary finance, but not at any cost whatsoever. There was a strong view that taking government money would be very damaging to the bank and the country, but 75% shareholder approval was required. There was also an element of personal interest amongst key players at Barclays. The bank had no bargaining position and the pool of investors was very small. Effectively, whatever the Qataris demanded, they got, including the $3bn loan and ASA 2 which gave them £280m over 5 years as a condition of the funding. Part of the reason why this was such an attractive deal was that the Warrants gave the prospect of large gains. Whilst the Warrants were in effect provided to the Subscribers “free of charge”, they were extremely valuable.

The RSAs and the RCI prospectus dated 25 November 2008 both stated that the issue price of the RCIs was 100%. There was no suggestion that only part of the amount paid for the RCIs would be eligible to qualify as Tier 1 capital.

The WSAs stated that the consideration for the issue of the Warrants was in each case 76p and the Warrant prospectus dated 25 November 2008 stated that the issue price was 0.01p per 100,000 Warrants.

We were taken to the terms of the RSAs and the WSAs. The RSAs provided that each Subscriber agreed to subscribe for £1.5bn in aggregate principal amount of RCIs and that the issue price of the RCIs was 100%. The obligation was conditional on shareholder approval. Shareholder approval was defined as approval of an increase in the authorised share capital of Barclays to enable the Warrants to be exercised and the MCNs to be converted.

The WSAs provided that the right to exercise the Warrants was subject to Barclays’ shareholders approving the issue of the Warrants and BBPLC issuing the RCIs and receiving full payment for the RCIs. In the event that those conditions were not met, then the Subscribers covenanted not to exercise the Warrants and there was provision for the Warrants to be transferred to a nominee of Barclays for a consideration of £1 and for the exercise price to be equal to the market value of the shares.

The FTT went on to make findings of fact under the following headings.

The FTT noted that there were no witnesses to explain the structure of the transactions, but it did have an internal paper setting out the expected accounting treatment of the RCIs. This was commenting on an earlier version of the transactions being negotiated which were very different to the actual transactions effected. At the time of the accounting paper it was contemplated that an issue of warrants would reduce the coupon on the RCIs by 2-3% to 10%. It noted that Barclays would not receive any immediate consideration but issuing warrants would reduce the RCI financing costs of BBPLC. There were practical reasons why Barclays could not issue RCIs and BBPLC could not issue warrants. The paper concluded that both the warrants and the RCIs should be recognised in the accounts at fair value and that the fair value of both was £3bn. An adjustment was necessary to allow Barclays to recognise the warrants at fair value and for BBPLC to recognise the RCIs at fair value.

As the majority of the cash would be received by BBPLC the relative transfer of value between the two should be treated as a capital contribution from Barclays to BBPLC. However, the FTT said that this appeared to be a conclusion of the accounting analysis. There was no evidence that the board of Barclays or BBPLC gave consideration to Barclays making a capital contribution to BBPLC. At an earlier stage of the negotiations when an issue of preference shares was being considered, it was noted that this would enable equity capital to be injected into BBPLC. The FTT noted at [100] that this was “some way from evidence that the Board of Barclays decided to make a capital contribution to BBPLC”.

Press comment at the time of the capital raising described Barclays shareholders as taking a £760m “hit” and the 14% coupon on the RCIs as a “nosebleed price”. Overall, the costs of the capital raising were described as a “thwack” to shareholders. It was noted that the market capitalisation of Barclays fell by £2.2bn after the deal was announced which suggested a discounting of the value transferred from existing to new shareholders. Part of the cost was said to be from “the huge pile of warrants Barclays has given the Abu Dhabi and Qatari Sheikhs”. Barclays’ share price fell from 205p to 179p on 31 October 2008. The Warrants were seen as a “one-way bet against existing shareholders”.

The FTT set out in detail a telephone conversation on 21 November 2008 between Mr Boath, the managing director of Barclays EU Financial Institutions Group and a senior figure in the capital raising, and Mr Stone, Barclays Group Treasurer.

The conversation concerned the accounting treatment of the Warrants and the RCIs. Our understanding of this conversation is that Mr Stone stated that the Warrants were being disclosed at £800m and the RCIs at £2.2bn. He noted the accountants were saying that they could not value the RCIs using the warrants first approach. His response was that the RCIs could not be valued separately as there was no market for them. The market was saying that if all the uncertainty about the capital raising plan was resolved then it would be looking for a coupon of 14% - 15%. Once all the equity was in place, investors bought the RCIs with a 14% coupon at par. The reality was that the Subscribers did not buy RCIs at par, they bought RCIs with a 14% coupon plus the Warrants. The real value of the RCIs was the amount raised after taking away the value of the Warrants. However, PWC were maintaining that it was necessary to look at the stand-alone value of the RCIs. Mr Stone described his conversations with PWC as bizarre.

A circular to shareholders issued on 7 November 2008 stated that the issue price of the RCIs was 100%. The FTT found that this would have been approved by the FSA.

An RCI prospectus was approved by the FSA as the UK listing authority on 25 November 2008. It noted that the RCIs would qualify as Tier 1 capital for the purposes of capital adequacy regulations.

The FTT did not describe the circular or the prospectus in detail and it did not expressly take them into account in its analysis of the transactions. However, we were taken to them in the course of oral submissions.

Notes to the unaudited pro forma financial information included in the prospectus stated:

The FTT noted that there was a regulatory requirement that Tier 1 capital must be fully paid up with the proceeds of issue being immediately and fully available to the firm. However, in correspondence between Barclays and the FSA in November and December 2008, Barclays described the RCIs as being shown in the accounts as having been issued for £2.2bn and the Warrants for £800m. This was said to be on the basis that it was necessary to establish fair values within the overall transaction price of £3bn.

The FSA challenged this analysis. It questioned why the value of £800m attributed to equity was being amortised if it was a capital contribution and stated that if it was not permanently available it should not be treated as Tier 1 capital. It would have expected Barclays to show the capital contribution in its accounts as a reduction in cash. The FSA indicated that its interpretation of IAS 32 (International Accounting Standard 32 – Financial Instruments: Presentation) was that the RCIs should be valued first when apportioning the £3bn and noted that a significant portion of the RCIs were sold to investors without warrants at the same coupon as those with warrants.

Barclays replied stating that the RCIs had been issued at a discount. Whilst the RCIs and the Warrants were issued by different entities, BBPLC received all the proceeds. It was management’s firm intention for Barclays to contribute the £800m payable for the Warrants to BBPLC as increased investment. The £2.2bn would be included in capital initially and that value would increase to par over the period to redemption, recognising that it was issued at a discount. Barclays noted that no RCIs were issued without warrants at the issue date. It said that a number of them were subsequently sold by the Subscribers weeks later in different market conditions. Barclays maintained that the FSA’s interpretation of IAS 32 only applied if a “compound instrument” was issued but that was not the case here.

The FTT found at [118] that there were no board minutes, internal papers or witness evidence to support the statement in the correspondence and in the RCI prospectus that Barclays intended to contribute £800m as a capital contribution to BBPLC. The suggestion of a capital contribution arose from an accounting analysis rather than an active decision of Barclays.

The FTT did not accept that none of the RCIs were issued by Barclays without warrants. It found that £500m of the RCIs were issued to the institutional investors without warrants.

The FTT noted that ultimately it appeared that the FSA agreed to follow the accounting treatment proposed by Barclays on the basis of Barclays’ interpretation of the apportionment of the £3bn. However, the FTT also noted that it had not been provided with evidence as to the conclusion of the correspondence between Barclays and the FSA. It considered that any agreement by the FSA as to the accounting treatment therefore added little weight to BBPLC’s case that its accounting treatment was correct.

We were told by Mr Prosser KC that the RCIs were innovative Tier 1 capital and that a contribution to equity would also be Tier 1 capital. However, the FTT does not appear to have placed any significance on the definition of Tier 1 capital. The FTT did note that the correspondence did not address the fully paid-up requirement, but it does not appear to place any significance on that requirement.

The FTT described the expert evidence before it at [122] – [127]. It noted that there were certain disagreements between the accounting experts. For example, the experts disagreed as to the date when the RSAs became loan commitments (whether that was 31 October 2008 when they were entered into or 24 November 2008 when shareholder approval was obtained) and the date when the RCIs should be valued (whether that was the commitment date or the issue date of 27 November 2008). The accounting experts also disagreed as to whether in substance the £3bn was paid for the RCIs only or for the RCIs and the Warrants, and as to whether it was IFRS-compliant to allocate the £3bn between the RCIs and the Warrants. However, the FTT considered that the question of whether the £3bn was paid only for the RCIs or for the RCIs and the Warrants was a matter of fact for it to decide.

The FTT also noted that there was disagreement between the valuation experts as to the fair value of the RCIs and the Warrants as at 31 October 2008 and 27 November 2008. However, the FTT considered that it would only need to engage with the valuation evidence if it decided that the £3bn was paid for both the RCIs and the Warrants. In the event it decided that the £3bn was paid only for the RCIs and therefore it did not make any findings on questions of valuation.

The FTT made what it described as “secondary findings” at [128] – [144]. In effect, these are the FTT’s principal findings on the question of whether the £3bn was paid only for the RCIs or for the RCIs and the Warrants. This is the most important part of the Decision for present purposes. We summarise the FTT’s secondary findings and its approach to the issues in the following paragraphs, including paragraph references to the Decision where appropriate.

The RCIs and the Warrants were negotiated as a package deal which also included MCNs, commissions and fees, ASA 2 and the unsecured loan. The Warrants were a very valuable part of the package deal. The experts valued the Warrants on 31 October 2008 at about £800m. The FTT described the issue as whether that value was in essence given away by the Barclays shareholders or whether the amount paid for the RCIs should be apportioned such that £800m was allocated to the Warrants ([129] and [131]).

The Warrants were clearly not going to be issued in isolation for a nominal consideration. However, the fact they had value did not in itself mean that Barclays would not give them away. It was necessary to look at the negotiations and the evidence as a whole ([132]).

There was a lack of evidence from witnesses who could have addressed the issue. The evidence before the FTT was limited to the documentary evidence and the judgment of Waksman J in PCP ([133]).

None of the documentation contemplated that the RCIs were being issued at anything other than the £3bn face value. The Warrants were not attached to the RCIs because they were issued by different companies. The internal accounting paper which suggested that RCIs might be issued below par was dealing with a different deal to that which was eventually negotiated. The Subscribers had been invited to invest £3bn for either RCIs with a higher coupon or RCIs with a lower coupon with warrants “attached”. In the event, the Subscribers insisted on a higher coupon together with the Warrants ([134] and [135]).

The FTT records at [136] a submission by Mr Prosser KC that the deal was a composite transaction as a matter of substance and economic reality. At [137], the FTT accepted that submission, in the sense that the Subscribers required both elements, but then explained how the composite transaction might be treated. It is worth setting out [136] and [137] in full:

The FTT then went on to conclude as a matter of fact that £3bn was paid for the RCIs and the value of the Warrants was given away by the Barclays shareholders. In reaching that conclusion it had regard in particular to various matters described at [138]. We shall deal with the sub-paragraphs in more detail when we come to address the criticisms made by BBPLC. At this stage we can summarise the matters relied on by the FTT at [138]:

The FTT went on to consider and reject various facts and arguments at [139] which it described as weighing against the conclusion it had reached at [138]. Again, at this stage we summarise the FTT’s approach and deal with the various sub-paragraphs in detail later in this decision:

Overall, these contrary arguments did not affect the FTT’s initial conclusion that in substance and as a matter of economic reality, the £3bn was paid for the RCIs ([140]).

The FTT then considered the effect of its findings of fact on the accounting issue at [141] – [144]. It noted at [141] that the opinion of BBPLC’s accounting expert on the basis of the facts as found was that the fair value of the RCIs at 31 October 2008, as defined in IAS 39 (International Accounting Standard 39: Financial Instruments: Recognition and Measurement), was the transaction price agreed between the parties. The transaction price was £3bn. The FTT rejected the opinion of HMRC’s expert that other valuation techniques should be used to establish fair value. However, the FTT noted that the opinion of HMRC’s accounting expert was that fair value of the RCIs at 27 November 2008 would be at least £3bn. It concluded that it did not need to address the valuation issue any further, or detailed arguments as to whether the accounting treatment depended on valuation at the commitment date or the issue date.

The FTT’s stated its overall conclusion at [161] as follows:

HMRC’s alternative case was that even if as a matter of substance and economic reality the £3bn was for the RCIs and the Warrants, the £800m did not fairly represent the losses of BBPLC arising from its loan relationships. As such, the effect of section 84(1)(a) and its successor provisions, until 2016 when they were repealed, was that no debit could be brought into account. This is what we identified above as Issue (4).

At first sight, it appeared to us that the FTT had dealt with this issue at [162] – [170] of the Decision and had accepted HMRC’s alternative case. If that is right, this appeal would be academic because there has been no appeal against that finding. However, the parties agreed matters were not that straightforward and that it was arguable the FTT was not addressing HMRC’s alternative case in these paragraphs. It may simply have been cross-checking its finding on the correct accounting treatment. If necessary, the FTT can clarify its finding on the alternative case in the event that the appeal is remitted to the FTT. In the circumstances, the parties did not make submissions on HMRC’s alternative case and we make no observations in relation to that aspect of the Decision.

There are two grounds of appeal for which the Upper Tribunal has given permission. Both grounds challenge the FTT’s finding that as a matter of substance and economic reality the Subscribers paid £3bn for the RCIs and the value of the Warrants was given away by the Barclays shareholders. BBPLC refers to this as “the Economic Reality Finding”. The grounds of appeal are as follows:

It is common ground that these challenges are challenges to the FTT’s evaluative finding of fact and fall within the principle in Edwards v Bairstow [1956] AC 1. We cannot set aside the FTT’s finding simply because we might disagree with it or consider it likely that we would have made a different finding if we had been determining the issue at first instance: see HMRC v Marlborough DP [2024] UKUT 98 (TCC) at [75]-[80].

BBPLC must establish an error of law. On Ground 1 it must show, in the words of Lord Radcliffe in Edwards v Bairstow,that “no person acting judicially and properly instructed as to the relevant law” could have concluded that the £3bn was paid for the RCIs. Lord Simonds described such circumstances as a “perverse” finding.

On Ground 2, BBPLC must show that the FTT took into account irrelevant considerations or failed to take into account relevant considerations. It must also show that the considerations wrongly taken into account or left out of account are material, in the sense that they might, not necessarily would, have affected the outcome: see Henderson LJ in Degorce v HMRC [2017] EWCA Civ 1427 at [95]. Where such a flaw in the fact-finding process is identified, there is no additional requirement to establish “perversity”: see Wm Morrison Supermarkets PLC v HMRC [2023] UKUT 20 (TCC) at [58].

Where an evaluative judgment is being challenged in the context of a multi-factorial evaluation of the evidence, it is well established that an appellate court or tribunal should exercise caution before interfering (see Fage UK Ltd v Chobani [2014] EWCA Civ 5 at [114]). Lewison LJ described the approach we should take in Volpi v Volpi [2022] EWCA Civ 464 at [2] and [5]:

We also take into account that we should read the FTT’s decision fairly and as a whole, without focusing too much on individual phrases or passages in isolation, and without being hypercritical (see DPP Law Ltd v Greenberg [2021] EWCA Civ 672).

HMRC’s case in relation to both grounds of appeal is that the FTT reached a conclusion which was not only open to it, but manifestly correct. It weighed the evidence and rightly concluded that the £3bn was paid for the RCIs with the Warrants being given away by the Barclays shareholders. There was ample evidence for this finding and the FTT considered all the contrary indicators.

There is considerable overlap in the two grounds of appeal. Essentially, we must consider under Ground 1 whether the FTT was reasonably entitled to reach a factual conclusion that the £3bn was paid by Subscribers for the RCIs with the value of the Warrants being given away by the Barclays shareholders. If the only reasonable conclusion on the basis of the FTT’s primary findings of fact was that £3bn was paid for the RCIs and the Warrants then we must allow the appeal and remit the appeal to the FTT to determine Issues (2) – (5).

Under Ground 2 we must consider whether the FTT took into account irrelevant factors or failed to take into account relevant factors which might have affected its overall evaluative conclusion. If it did, we must consider whether those factors were material to its conclusion and if so, either re-make the decision or remit it to the FTT for it to re-make the decision.

The submissions of Mr Prosser KC and Mr Henderson for BBPLC were that the FTT reached a conclusion that was not reasonably open to it for the following broad reasons:

It is common ground that the FTT was required to consider the nature of the transactions in accordance with their substance and economic reality. Ms Wilson KC and Mr Simpson for HMRC submit that the FTT was entitled to come to the conclusion it did as a matter of substance and economic reality. It was entitled to take into account the particular factors it described at [138] and [139] in reaching that conclusion.

As noted above, the FTT addressed two analogies relied on by BBPLC. We consider the FTT’s approach to those analogies below. At this stage we can address another argument by analogy relied on by BBPLC before us. The following fact pattern was put forward: A agrees to provide two items to B; A and B know that both items are very valuable; B agrees to pay a substantial sum to A; A would not agree to provide either item in an isolated transaction for free and only agrees to provide them both if B pays the full amount; B would not agree to pay the sum for only one of the items. BBPLC submits that in these circumstances the only possible conclusion is that the substance and economic reality of the transaction is that it is a package deal where the sum is paid for both items. That is the case irrespective of the legal form of the transaction.

We do not consider that this analogy assists BBPLC’s case. It is simply a description of the fact pattern in the present case devoid of any context. In our view, whenever the substance and economic reality of a transaction falls to be considered, it is always necessary to consider the transaction in context. We agree with Ms Wilson KC there are many kinds of package deal and if it is necessary to identify the substance and economic reality of the deal then it must be considered in context.

It appeared to be common ground that the substance and economic reality of a transaction is to be determined objectively. The subjective beliefs of the parties to a transaction will not be directly relevant. However, we agree with Ms Wilson KC that views expressed contemporaneously might act as a legitimate cross-check on what is an objective question. If the view of a tribunal is that a transaction in substance falls to be described in a particular way, the tribunal is entitled to test that view against the views of others who might have expressed a different view at the time of the transaction.

BBPLC submits that as a matter of substance and economic reality, and taking into account the following findings of fact, the only possible conclusion open to the FTT was that the £3bn was paid for the RCIs and the Warrants:

BBPLC submits that the FTT’s analysis at [138] and [139] is a perverse economic analysis. The various factors relied on by the FTT could not alone or together justify that conclusion, and some of those factors “defy common sense”. The FTT recognised that the deal was conditional at [137] but failed to recognise that the conditionality was a factor supporting BBPLC’s case.

It does not appear to us that the FTT did overlook that the Warrants were not exercisable by the Subscribers unless and until the £3bn was paid in full or that the Warrants were liable to forfeiture if the £3bn was not paid by the due date. The FTT made a primary finding of fact at [28(1)(b)] that the right to exercise the Warrants was conditional on BBPLC issuing the RCIs and receiving full payment in respect of the RCIs.

The FTT reached its overall conclusion at [140] that the £3bn was paid for the RCIs alone on the basis of all its secondary findings at [128] – [139]. It is true that the FTT described particular factors to which it had regard at [138] and then at [139] it considered particular facts and matters weighing against what might be described as its initial conclusion. However, we must have regard to the secondary findings as a whole. Immediately prior to [138], at [136] and [137], the FTT made findings in relation to the conditionality of the RCIs and the Warrants. Further, the FTT had conditionality in mind when it stated at [138(4)]:

BBPLC submits that if the FTT did take into account this conditionality, then it failed to give reasons why it did not weigh in favour of the £3bn being paid for the RCIs and the Warrants. We do not accept that submission. It was plainly a relevant factor and it would have been better if the FTT had expressly identified it as such at [138] or [139]. However, having found that the FTT took this factor into account, the question of weight was a matter for the FTT. It is sufficient that the FTT took it into account in its evaluative judgment.

HMRC say that the six factors identified by BBPLC are not the only relevant factors. There are other factors which are relevant in determining the substance and economic reality of the transactions. In particular:

Factors (4) – (6) were expressly dealt with by the FTT at [138] and [139] and we shall consider their relevance when we come to look at Mr Prosser KC’s specific criticisms of the FTT’s reasoning in those paragraphs. At this stage we consider factors (1) – (3).

The first factor relied upon by HMRC refers to the terms of the legal documentation. The FTT took as its starting point the contractual terms of the RSAs and the WSAs, stating as follows at [138(1)] and [139(2)]:

Both parties agree that the FTT was correct to take the contractual terms as its starting point.

HMRC submits that the contractual terms are strong contemporaneous evidence of the common intention of the parties. They are a good way of describing what the FTT eventually decided was the substance of the deal and they make commercial sense. Barclays was trying to raise regulatory capital by way of the RCIs. The Warrants were a “sweetener”. The Subscribers would not get their sweetener unless they subscribed for and paid for £3bn of RCIs. The circular and the announcement make no suggestion that this is not in substance £3bn for the RCIs. It was how the deal was described to the world. Even if the documents are open to two interpretations as a matter of substance, then the FTT was entitled to find that the substance was that the £3bn was paid for the RCIs alone. In those circumstances, Ground 1 must fail.

HMRC also submitted that the RCIs being issued at par was consistent with the FSA’s requirement that to qualify as Tier 1 capital the RCIs had to be fully paid up. However, we were not taken to any definition of what amounts to Tier 1 capital, or indeed what amounts to “fully paid up” in this context and it was not addressed in the FTT’s decision. It is also the case, as BBPLC submits, that the FTT did not rely on any regulatory requirements to justify its conclusion.

The second and third factors relied on by HMRC are not expressly taken into account at [138] and [139] or discussed at [128] – [144]. It is therefore difficult to discern from the FTT’s secondary findings whether it took those factors into account. However, in our view the fact that BBPLC was beneficially entitled to the full £3bn and the absence of a firm intention on the part of Barclays to make a capital contribution of £800m to BBPLC were relevant factors in assessing the substance and economic reality of the transactions. The economic reality has to reflect why the £800m did not go to Barclays. BBPLC was trying to explain that fact by inviting a finding that Barclays had a firm intention to make a capital contribution to BBPLC but the FTT declined to make that finding.

The first suggestion of a capital contribution comes in the internal accounting paper which was undated but was prepared prior to 31 October 2008. The FTT held that the reference to a capital contribution was a conclusion of the accounting analysis and not evidence of a decision to make a capital contribution. There was then a reference in Note 5 to the unaudited pro forma financial information included in the RCI Prospectus dated 25 November 2008 quoted above, which also identified in Note 6 for the purpose of “clarification” that at the time the deal was announced the proceeds were treated as liabilities and not as equity capital. It was also discussed in the correspondence between Barclays and the FSA in November and December 2008. There was no evidence that at 31 October 2008 the board of Barclays or BBPLC had agreed or firmly intended that Barclays would make a capital contribution of £800m to BBPLC. The public announcement on 31 October 2008 stated that the objective of the plan to raise Tier 1 capital would be met through the issue of the RCIs.

BBPLC says that the terms of the deal between BBPLC and Barclays are irrelevant to the question of what the Subscribers are paying for. It is irrelevant that two different companies are issuing different instruments or that one of the companies is getting all the money. We do not agree. In our view, the agreement between Barclays and BBPLC is relevant in identifying the substance of the agreement as between Barclays and BBPLC on the one hand and the Subscribers on the other.

We now turn to consider BBPLC’s eleven specific criticisms of the FTT’s reasoning at [138] and [139]. BBPLC submits that there are eleven factors which the FTT wrongly took into account in determining the substance and economic reality of the transactions. We deal with those factors adopting the same numbering as the skeleton argument and Mr Prosser KC’s oral submissions. BBPLC says that these factors which were taken into account by the FTT could not possibly justify the FTT’s conclusion and some of them defy common sense.

The institutional investors subscribed for £500m RCIs at par

The FTT stated at [138(2)]:

BBPLC submits that these facts are not relevant to the issue, which concerns only the transactions agreed by BBPLC, Barclays and the Subscribers. The agreement between the institutional investors and BBPLC was a separate deal, between different parties at a later date in very different economic circumstances. By 19 November 2008, when it was announced that the institutional investors had fully taken up the £500m of RCIs made available to them, Barclays had already secured £7.3bn of additional capital with the result that the perceived value of the RCIs had changed. This transaction says nothing about the substance and economic reality of the earlier transactions. The Subscribers agreed to make £500m of RCIs available to institutional investors but remained liable to subscribe for them if they were not taken up.

HMRC submits that whilst the institutional investors did not get any of the Warrants, they paid par value for the exact same RCIs. The FTT was entitled to take that into account. The amended deal is consistent with, and supportive of, the FTT’s conclusion. The Warrants were untouched, having done their separate job of getting the Subscribers to invest £3bn for the RCIs. BBPLC had invited the FTT to make a finding that the deal with institutional investors was a different deal, between different parties at a later date in very different economic circumstances, but the FTT did not make that finding.

In the course of submissions, we were taken to the “concession letters” signed by the Subscribers on 18 November 2008. The Subscribers confirmed that they would each make available up to £250m of the principal amount of the RCIs. It was stated, for the avoidance of doubt, that the Subscribers were not transferring any Warrants. The RCIs subscribed for by the institutional investors were to be issued directly by BBPLC to the institutional investors. The Subscribers remained obliged to subscribe and pay for any RCIs not taken up by the institutional investors.

The deal with the Subscribers needed shareholder approval and this offering to the institutional investors was made in order to obtain that approval. To that extent, we consider that the FTT was entitled to view it as part of the original transactions, although it was not specifically contemplated at the time the RSAs were entered into on 31 October 2008. Incidentally, the same approach might be taken to the possibility of a capital contribution which was not identified until 25 November 2008. The issue of RCIs at par to the institutional investors can also be viewed as being carved out of the deal to issue the RCIs at par to the Subscribers. In that sense it can be said that there is consistency between these aspects of the transactions. It also illustrates that the Warrants were not attached to the RCIs. The FTT was entitled to give some weight to the fact that the institutional investors got the same deal, but without the Warrants. In giving weight to this factor, the FTT took into account at [138(2)] that the perceived value of the RCIs had changed since 31 October 2008. It was a matter for the FTT to determine how much weight to give to this factor. We do not consider that this factor had no relevance or that the FTT was not entitled to give it any weight.

HMRC also submitted that the Subscribers were not going to give up the sweetener of the Warrants just because they were releasing some of the RCIs. That is because the sweetener did a different job. It got the deal away. We deal with HMRC’s description of the Warrants as a “sweetener” below.

Press comment at the time

The FTT said as follows at [138(3)]:

BBPLC submits that this was irrelevant because the various commentators were not considering the same issue as the FTT. They were criticising Barclays for agreeing what they perceived to be, in the round, an expensive fundraising exercise. The shareholders would have taken the same “thwack” and similar criticism would have been made if the legal form of the deal had been that the Subscribers were paying £2.2bn for the RCIs and £800m for the Warrants. The substance and economic reality of the deal had no effect on the criticisms being made. The effect on shareholders would be the same however the deal was described.

HMRC submits that the view of external commentators is corroborative of the FTT’s finding as to the substance and economic reality of the transactions. These were contemporaneous industry views, consistent with what Barclays was announcing publicly and based on objective information released by Barclays in relation to the capital raising. The FTT was entitled to take those views into account. The analysis was not just that the deal was expensive for Barclays, but that the shareholders were taking a “thwack”.

The FTT appears to have been considering at [138(3)] whether there was anything to suggest that the legal form of the transactions, that the Warrants were being issued for a nominal consideration, did not reflect the substance of the transactions. It considered two aspects of the evidence in this regard. The views of external commentators, which is the subject of this criticism, and the findings of Waksman J in PCP, which is the next criticism.

We consider that the contemporaneous views of external commentators may be a relevant consideration, but only as a cross-check against a conclusion reached based on the objective evidence. The external commentators are construing the same publicly available material which was available to the FTT. However, even on a generous reading of the Decision it is not apparent to us that the FTT was using this evidence as a cross-check to its conclusion. Rather, it appears to be evidence which the FTT took into account in reaching its conclusion. This is therefore an error of law on the part of the FTT.

There is also merit in Mr Prosser KC’s submission that in any event the external commentators were not considering the same issue. They were considering the effect of the transactions on the Barclays shareholders. That was the context in which it was said that the shareholders were giving away £800m of value. They were not considering the substance and economic reality of the transactions. We agree that such views should carry no weight in determining the substance and reality of the transactions. Even if the FTT had been using this material as a cross-check, it would have provided little if any support for its conclusion.

Issuing the Warrants cost Barclays nothing

The FTT went on to say as follows in [138](3)]:

The FTT makes the same point about the Warrants being given away in various other paragraphs including at [139(1)]) where it stated as follows:

The description of value being given away by shareholders rather than by Barclays was one reason why the FTT rejected a submission by BBPLC that the RCIs and the Warrants were part of a package for which £3bn was paid and distinguished the car showroom example at [139(5)]:

Mr Prosser KC submitted that the car showroom analogy is clearly analogous to the present facts. However, the FTT did not explain why the difference between Barclays giving away value and Barclays’ shareholders giving away value was so fundamental. Why should it matter who is giving away value? In any event, Barclays was the one issuing the Warrants. The shareholders were suffering a loss, but it was a reflective loss.

Mr Prosser KC submitted that the FTT’s reasoning defies commercial common sense. It was Barclays as the issuer of the Warrants, and not its shareholders, which could require payment of £800 million from the Subscribers in return for issuing them. If, as a matter of substance and economic reality, the £3 billion was for the RCIs alone, it was self-evidently Barclays which gave away £800 million of value in omitting to obtain that amount for the Warrants.

We do not consider that Mr Prosser KC’s analogy to a car showroom is particularly helpful. As Ms Wilson KC submitted, cars are not debts and car servicing is not equity. As we have said above, the context in which the issue arises is relevant and important. We are concerned with the accounting treatment of specific transactions, intended to raise regulatory capital, more particularly the substance and economic reality of those transactions.

We do however accept Mr Prosser KC’s submission that it is irrelevant whether the transaction is viewed as Barclays giving away value or the Barclays’ shareholders giving away value. The FTT appears to consider that this was a fundamental factor in its analysis at [138(3)] and [139(5)]. We cannot see that this was a relevant factor in determining the substance and economic reality of the transactions, and whether the Warrants were being given away as a sweetener to get the deal done for the RCIs. In our view, it certainly was not a fundamental factor.

We were told that the FTT had been invited to find that Barclays would not have been willing to give the Warrants away. The FTT at [132] accepted that Barclays would not issue the Warrants in isolation for nominal consideration but it did not find that Barclays would not give the Warrants away in the context of an overall package.HMRC therefore says that the FTT was entitled to find that the Warrants were being given away in the sense of a sweetener to get the deal done. We consider that question below. At this stage, we are satisfied that the FTT fell into error in [138(3)] in treating as relevant that it was Barclays’ shareholders giving away value rather than Barclays itself.

Misplaced reliance on PCP

BBPLC submits that the FTT also erred at [138(3)] in wrongly relying on an observation by Waksman J in PCP that “the Warrants had been given away”.

We agree that this observation was not relevant to the issue before the FTT. The FTT itself noted that Waksman J was not considering the economic substance of the deal. It justified relying on the observation on the basis that it was made at the same time as recognising that the Warrants were extremely valuable. We cannot see that this context makes the observation relevant to an analysis of the substance and economic reality of the transactions.

At best, the observation of Waksman J might have been used as a cross-check to a conclusion reached on the basis of the evidence before the FTT. We acknowledge that the FTT had no evidence from individuals involved in the transactions whereas Waksman J did have such evidence. However, that does not justify the FTT placing reliance on his observation. It is true that Waksman J gave an impressively forensic judgment. But he was not considering the substance and economic reality of the transactions for accounting purposes. The issue before him was whether Qatar had got a better deal than PCP in circumstances where Barclays had represented that they would get the same deal. In relation to the Warrants, Qatar and PCP got the same deal.

It does not appear that the FTT was using the observations of Waksman J as a cross-check to its conclusion. If anything, it was using those observations as a cross check to the views of external commentators and found that the observations were consistent with those view. As such, we are satisfied that the FTT did err in law at [138(3)] in placing reliance on the observations of Waksman J in PCP.

Subjective perceptions

We agree with Mr Prosser KC that identifying the substance and economic reality of the transactions requires an objective approach. Ms Wilson KC did not seriously contend otherwise. An entity is required to account for transactions in accordance with their substance and economic reality, not in accordance with the subjective perceptions of its directors, employees or others.

Mr Prosser KC submitted that the FTT wrongly took into account subjective perceptions in reaching its conclusion. This was a new argument which did not form part of BBPLC’s original grounds of appeal. However HMRC took no objection to the point being taken.

The FTT stated at [138(3)] that the judgment of Waksman J does not suggest that anyone regarded the RCIs as having been issued other than at par. That is not surprising because it was not relevant to the issues he had to decide. The FTT should not therefore have given it any weight. If there had been such a suggestion, then in our view the FTT may have been entitled to take it into account as a cross-check depending on the circumstances in which it arose.

The error is also said to be apparent in various other passages in the Decision. In particular:

Further, at [139(3)] the FTT referred to “what was being considered”:

It is said that overall the FTT placed considerable weight on subjective perceptions in reaching its conclusion on the substance and economic reality of the transactions, and the absence of any evidence of those subjective perceptions. No evidence was required because the FTT had all it needed to know in the six factors identified by BBPLC as being relevant. Mr Prosser KC submitted that everything else was “noise”.

We do not accept these submissions, or that the FTT was wrong to note the absence of such evidence. Essentially, the FTT was looking to see if there were any clues as to the substance and economic reality of the transactions in the available material. Witnesses may have been in a position to provide context to the deal and the negotiations in addition to that appearing in the documents. The FTT might discount subjective views entirely, or it might test its conclusion for coherence in the light of those views. Looking at the Decision fairly and as a whole, we do not consider that the FTT relied on subjective perceptions in reaching its conclusion. The FTT was asking itself whether there was anything to suggest that it might be wrong in its conclusion. If the evidence appears to point to a particular conclusion, but someone has a different understanding, it is entirely appropriate for a tribunal to pause for thought and seek to reconcile the evidence if possible.

On that basis, the FTT was entitled to note and indeed draw comfort from the fact that the judgment in PCP did not suggest that anyone regarded the RCIs as having been issued other than at par which was consistent with its own views. It might have made it clearer that that is what it was doing, but overall we do not consider that this criticism of the Decision is justified.

It is also worth noting at this stage that the way in which the transactions were presented to Barclays’ shareholders and markets generally in the circular to shareholders and in the prospectuses would be relevant evidence which the FTT was entitled to take into account. They are not simply subjective views from within Barclays.

The public announcement on 31 October 2008 and the RCI prospectus were consistent with the contractual documentation in stating that the Warrants had been issued for nominal consideration and the RCIs had been issued at par. The FTT was also entitled to observe at [134]:

The circular issued to Barclays shareholders on 7 November 2008 ahead of a general meeting on 24 November 2008 described how Barclays was raising more than £7bn of additional capital, including an issue of £3bn of RCIs. It also stated that “in conjunction with this issue, [the Subscribers] have also subscribed (for a nominal consideration) for warrants …”. Under the heading “Issue price”, the circular states “100% (£50,000 per RCI)”.

The FTT did not expressly rely on the circular in its secondary findings, but in this section of the Decision it clearly had in mind all the documentation identified earlier in the Decision. That included the RSAs, WSAs, the circular and the prospectuses.

The FTT was also entitled to discount at [139(6)] the subjective views of Mr Stone and Mr Boath in their telephone conversation. There was no criticism of the FTT’s approach to this aspect of the evidence.

Qatar’s demands

The FTT said as follows in [138](4):

BBPLC makes several criticisms of this paragraph. Under this heading, it is said that the FTT’s reliance on findings made by Waksman J in PCP that Qatar got what it wanted and that these were extraordinary times were not relevant.

The FTT had already noted at [90(1)-(3)] various findings by Waksman J in PCP. Barclays was desperate, in the sense that it would go to extraordinary and unusual lengths to bring about the capital raising, although not at any cost whatsoever. No bank had any bargaining position of strength in October 2008 and the pool of investors was extremely small. Effectively, the Qataris got whatever they demanded in terms of benefit.

Nevertheless, the FTT found at [68] that the transaction was negotiated on arm’s length terms and at [69] that Barclays was not forced to deal with the Subscribers. BBPLC submits that in the light of those findings, Barclays’ “desperation” and the fact that the capital raising was agreed in “extraordinary times”, were not relevant to the issue. They tell us nothing about what the £3 billion was for as a matter of substance and economic reality.

We do not accept that those findings were irrelevant. The FTT had to consider the specific transactions and the circumstances in which they came about, including how the negotiations progressed. Qatar wanted another £400m of value to do the deal, and they got it. BBPLC and Barclays changed position from offering the RCIs at 14% with no Warrants, or the RCIs at 10% with Warrants, to the RCIs at 14% and the Warrants. The explanation for that, at least in part, was Qatar’s hard bargaining in extraordinary times.

Other valuable elements of the deal demanded by Qatar

The second criticism of [138(4)] is that the FTT regarded the fact that the Warrants, the unsecured loan and the various fees were all elements demanded by Qatar as supporting its view of the substance and economic reality of the transaction. The FTT also refers at [137] to other elements of the deal giving value to Qatar:

The FTT noted at [130] that BBPLC was not arguing that the £3bn was also paid for these elements of the deal demanded by Qatar apart from the Warrants. BBPLC submitted that the FTT wrongly considered that this undermined its arguments. The fact that there were other elements to the package deal which BBPLC was not arguing were part of what the £3bn was paid for, should not have been a reason for the FTT to conclude that the £3bn was not for the Warrants.

In our view it is not fair to suggest that the FTT regarded the existence of these other benefits as inconsistent with and thereby weakening BBPLC’s case that the £3bn was for the RCIs and the Warrants. Read as a whole, the FTT was simply describing the context of the transactions. However, we do not accept a submission by HMRC that the fact the RCIs and the Warrants were part of a wider package deal of benefits which were not reflected in any discount in the RCIs supported its view that the RCIs should not be discounted at all in relation to the Warrants.

The Warrants as “sweeteners”

The third criticism of [138(4)] relates to the FTT’s acceptance of Ms Wilson KC’s submission that if the Warrants were valued at £800m then they could not be a sweetener for the deal because that was their actual value. BBPLC says that the FTT’s observation defies common sense. The Subscribers were getting a good deal because they got the RCIs and the Warrants for £3bn. The Warrants were a sweetener because Barclays had to provide them to persuade the Subscribers to participate in the fundraising. The Subscribers got exactly the same good deal, and the Warrants were still a ‘sweetener’, whatever the substance and economic reality of the transactions. The allocation of the £3bn between the RCIs and the Warrants does not affect the substance and economic reality of the transaction. They could operate as a sweetener without being given away. For example, Waksman J at [497] described the $3bn loan as being “part of the price for the deal”. The Warrants were similarly part of the price for the deal.

Mr Prosser KC acknowledged that we do not know whether £800m is properly attributable to the Warrants because the FTT has not yet decided whether the “warrants first” approach is appropriate.

HMRC submits that the concept of a sweetener to a deal does not defy common sense, and it has a commercial logic in the present context. It fits the facts of the case. The Warrants could be used in different ways. As a proxy for interest or being given away as a sweetener to motivate the Subscribers to do a deal for the RCIs. They were motivational because they were priced at £1.52. If it is possible to look at the deal in different ways then the FTT was entitled to make an evaluative judgment as to the substance and economic reality of the deal. The deal was that Barclays needed £3bn of regulatory capital and the Warrants enabled Barclays to obtain that regulatory capital.

It does appear to us that in [138(4)], the FTT wrongly considered that the Warrants would not be a sweetener unless they were given away. The Subscribers would be getting a good deal whether or not part of the £3bn was attributed to the Warrants. The Warrants could still be viewed as a sweetener. In the FTT’s words, the Subscribers would obtain “very real value out of the deal” because they were getting the RCIs and the Warrants. The Warrants were motivational because they were included in the deal without increasing the £3bn which was to be paid. It does not necessarily follow, as the FTT appears to have thought, that the £3bn was being paid for the RCIs and the Warrants were only motivational because they were being given away.

In one sense this criticism depends on how “the deal” is defined. If it is a deal for £3bn of RCIs then the Warrants would be a sweetener to that deal if they were given away. If however, the deal was to raise £3bn of Tier 1 capital and this could only be done by offering RCIs and Warrants then the Warrants were not a sweetener as such. They were simply part of the deal to raise £3bn. HMRC’s case is essentially that what Barclays wanted was £3bn of regulatory capital in the form of the RCIs. The Warrants enabled that deal to go ahead. BBPLC’s case is essentially that the deal was £3bn for the RCIs and the Warrants and that it was a good deal for the Subscribers. If that is right, then the Warrants would not be a sweetener as such.

Either way, we are satisfied that the FTT’s approach to the Warrants in this paragraph does amount to an error of law. It was not the case that the Warrants could only be a sweetener if they were given away. It appears to us that the FTT’s approach begs the question as to the nature of the deal. It assumed the deal was for the RCIs with the Warrants being given away as a sweetener.

The absence of any reduction in the RCI coupon

The fourth criticism of [138(4)] is the FTT’s reliance on the fact that the Subscribers obtained the Warrants without any compensating reduction in the RCI coupon. The FTT also said at [138(5)]:

BBPLC submits that the absence of any reduction in the coupon is not relevant to substance and economic reality. That is because the fact that the Subscribers succeeded in persuading Barclays to provide the Warrants as well as the RCIs, without having to make any concession of their own regarding the coupon on the RCIs, tells us nothing about what the £3bn was for as a matter of economic reality.

We do not accept that submission. In our view the FTT was entitled to take into account that at one stage of the negotiations, Barclays was offering a coupon of 14% or a coupon of 10% plus the Warrants. It is clearly relevant in those circumstances that the Warrants are serving as a proxy for interest, and that value is being attributed to the Warrants. The counter-offer by Qatar is a coupon of 14% and the Warrants for their £1.5bn. The FTT was entitled to consider that this was a relevant factor which reduced the strength of any suggestion that in substance the Subscribers were paying for the Warrants by reference to a reduction in the coupon.

Approval of the accounts by PWC

The FTT said as follows at [139(4)]:

It seems clear from this paragraph that the FTT placed no weight on the fact that PWC had approved accounts showing that the RCIs were issued at a discount and, by inference, that £800m was paid for the Warrants which was treated in the accounts as a capital contribution by Barclays to BBPLC. The FTT also placed little weight on the expert accounting evidence before it. There is no challenge to the FTT’s treatment of the expert accounting evidence on this issue. However, BBPLC submits that the FTT should have regarded PWC’s approval of the accounts as at least an indicator that they were IFRS-compliant in accounting for the £3bn as paid for the RCIs and the Warrants as a matter of substance and economic reality. It also appears that the reason the FTT placed no weight on PWC’s approval of the accounts was the absence of evidence from PWC to explain the basis for its conclusion.

We do not consider that this criticism establishes any error of law. The FTT was entitled to determine the substance and economic reality of the transactions by reference to the direct evidence before it as to the terms of the transactions and the circumstances in which the transactions took place. If the FTT was entitled to give little or no weight to the expert evidence on that question, which was common ground, it was entitled to treat PWC’s audit opinion in the same way, especially in the absence of any evidence from PWC and any opportunity to test how PWC had concluded that BBPLC had adopted an appropriate accounting treatment.

The FTT refers to the fact that PWC were persuaded to change their view on the “warrants first” approach to valuation of the RCIs. That is a different issue and on the face of it would not be relevant in determining the substance and economic reality of the transactions. Nevertheless, we regard this as no more than a loose reference by the FTT, rather than a particular reason why it was discounting PWC’s audit report on Issue (1), which it was dealing with at [139]. It is tolerably clear that the FTT wanted to know what had persuaded PWC to accept the accounting treatment, not just the valuation. That would have been factual evidence, not expert evidence of opinion. It would also have been evidence the FTT could have legitimately taken into account in considering the substance and economic reality of the transactions.

A differently constituted FTT might have given the audit report some weight. However, we do not consider that the FTT can be criticised for not giving weight to the fact that PWC had approved the accounts.

Relevance of Marks and Spencer

The FTT considered Marks and Spencer at [139(7)] and concluded that it had little bearing on the issue it had to determine:

BBPLC submits that the FTT had no good reason to distinguish Marks and Spencer. The concept of economic and commercial reality considered by the FTT and Upper Tribunal in that case did not have a specific EU or VAT meaning. The consequence of having untaxed consumption of wine in circumstances where the taxpayer had obtained input tax credit on purchase of the wine was a completely separate “cross check” applied by the FTT and the Upper Tribunal in respect of the VAT conclusion they had reached. It was not a reason given in support of their economic and commercial reality analysis of the package deal. It was only after the Upper Tribunal had carried out its analysis based on VAT law that it went on to consider the economic and commercial reality of the supplies.

In our view it was not necessary for the FTT to distinguish the Upper Tribunal decision in Marks and Spencer. The substance and economic reality of a transaction is an evaluative judgment which is particularly fact sensitive. Different tribunals may legitimately reach different conclusions on that question, even on the same facts. As we have said, it is necessary to consider the nature of the transactions in their context. It is not possible to say that in any package deal, the economic reality will always be that what is paid for the package is paid for all elements of the package. The context, including the negotiations, may indicate that one element of the package has been given away, for example as a sweetener.

What is apparent from [139(7)] is that the FTT again considered the fact that value was moving from Barclays’ shareholders rather than from Barclays itself was an important factor in analysing the substance and economic reality of the transactions. For the reasons given above, we consider that was an error of law.

For the reasons given above, we do not accept Mr Prosser KC’s submission that the only conclusion available to the FTT was that the £3bn was paid for the RCIs and the Warrants. The FTT was right to consider in some detail the broader context in which the transactions took place. We do not accept that the six factors identified by Mr Prosser KC are the only relevant factors and lead inevitably to that conclusion.

The FTT conducted the evaluative exercise that it was required to do. Whilst the FTT was wrong to rely on some of the matters which Mr Prosser KC has identified, it was entitled to rely on the other matters. Overall, it cannot be said that there was only one conclusion available to the FTT.

We have dealt under Ground 1 with all the factors which BBPLC says the FTT should not have taken into account because they were irrelevant. For the reasons given above, we are satisfied that the FTT did err in its analysis in the following respects:

We are satisfied that if the FTT had not erred in these respects, then it might have reached a different decision on the substance and economic reality of the transactions. In those circumstances, we allow the appeal on Ground 2 and set aside the Decision of the FTT.

The parties took different positions as to what we should do if we allowed the appeal on Ground 2. Mr Prosser KC originally submitted in his skeleton argument that we should remit the matter to the FTT with directions for it to reconsider its decision in the light of our decision. However, in oral submissions Mr Prosser KC invited us to re-make the decision if we felt able to do so. Ms Wilson KC invited us to either re-make the decision or remit as we might think fit.

We are conscious that neither party seeks to go behind the FTT’s findings of primary fact, or invites us to make any additional findings of primary fact. We have been taken to some of the relevant documentary evidence in the course of submissions. However, after careful consideration, we consider that the FTT would be in a better position to re-make the decision. It has had the benefit of hearing full submissions on the documentary evidence as a whole. It may wish to hear further submissions in the light of this decision. Having regard to the overriding objective we consider that the FTT should re-make the decision on Issue (1) (see [13] above for the Issues). Irrespective of how the FTT re-makes the decision on Issue (1), we consider that it should also clarify its findings on Issue (4) and make findings on issues (2), (3) and (5).

We should also note that BBPLC relied on what it says are two relevant factors which the FTT failed to take into account and which would have been material to the outcome.

Firstly, the FTT found at [28(1)(b)] that the terms of the WSAs and the Warrants were such that the Warrants were not exercisable unless and until the £3bn was paid in full. Indeed, the Warrants could be forfeited if the £3bn was not paid. BBPLC submits that this was highly relevant to the substance and economic reality of the transaction but that the FTT failed to take it into account.

For the reasons given above, we are satisfied that this is a relevant factor but that the FTT took it into account, although it is unclear the weight that the FTT put on it as it was not identified as a particular factor in [138].

Secondly, BBPLC submitted that if subjective perceptions are relevant, the FTT failed to take into account various board minutes and other documents in which the Warrants were described as “attached” to the RCIs and as being “detachable”. Barclays was discussing with the Subscribers a subscription for RCIs which would carry a 10% coupon and have detachable warrants. We have found that subjective perceptions were only relevant as a cross-check.

For the reasons given above, we allow the appeal on Ground 2. We remit the appeal to the FTT to reconsider its decision on Issue (1) in accordance with our findings in this decision, and to determine Issues (2) – (5).