BNP Paribas SA (London Branch) v Revenue and Customs (INCOME TAX/CORPORATION TAX : Anti-avoidance) [2017] UKFTT 487 (TC)

FTT-Tax

BNP Paribas SA (London Branch) v Revenue and Customs (INCOME TAX/CORPORATION TAX : Anti-avoidance)

[2017] UKFTT 487 (TC) · 2016-06-02

[2017] UKFTT 487 (TC)

[image removed] TC05941 Appeal number: TC/2011/02091 CORPORATION TAX - whether price of purchase of right to dividends deductible - whether purchase and sale of right to dividends was trading transaction in course of Appellant’s trade - no - whether purchase price expenditure incurred wholly and exclusively for purposes of the trade - no - whether HMRC permitted to argue point in relation to section 730 ICTA that was not raised in closure notice and which they stated they were not pursuing - yes - whether price of sale of right to dividends disregarded for purposes of calculating Appellant’s trading profits under section 730(3) ICTA - no - appeal dismissed FIRST-TIER TRIBUNAL TAX CHAMBER BNP PARIBAS SA (LONDON BRANCH) Appellant - and - THE COMMISSIONERS FOR HER MAJESTY’S Respondents REVENUE & CUSTOMS TRIBUNAL: JUDGE GREG SINFIELD JUDGE HARRIET MORGAN Sitting in public at the Royal Courts of Justice, Strand, London WC2 on 23 to 27 May and 31 May to 2 June 2016 Michael Flesch QC, Richard Boulton QC and Michael Jones, counsel, instructed by Clifford Chance LLP, solicitors, for the Appellant Giles Goodfellow QC and David Yates, counsel, instructed by the General Counsel and Solicitor to HM Revenue and Customs, for the Respondents © CROWN COPYRIGHT 2017 DECISION

[1]The appeal is against additional corporation tax which HMRC seek to impose in relation to the purchase and sale of a right to certain dividends which BNP Paribas SA (“BNP”) undertook through its London branch (the “London branch”) in the accounting period ending 31 December 2005.

[2]In short, the London branch submitted its corporation tax return for that period on the basis that the transaction generated a trading loss of £96,091,000. This loss arose, in its view, because the price paid for the right to the dividends was tax deductible from the profits of its banking trade but the sale price received was not a taxable receipt of that trade due to the operation of s 730(3) of the Income and Corporation Taxes Act 1988 (“ICTA”). HMRC disagreed with both of these propositions. Background

[3]There are so many corporate entities and individuals referred to in this decision that, for ease of reference, we provide a list of the main characters and abbreviations used in this decision in an Appendix to this decision. Outline of the facts

[4]The facts are set out more fully below but, in summary, the key transactions are as follows:(1) On 13 December 2005, the dividend rights were created by a new company, Harewood Investments No.5 Limited (“HIL”), issuing 1.5 million ordinary shares (the “shares”) to a subsidiary of BNP, BNP Luxembourg SA (“BNP Lux”), for a total subscription price of £210 million. The shares carried the right to a dividend payable monthly at an effective fixed rate of 4.354% on an amount of £150 million and to a termination dividend of a total of £150 million payable no later than 15 December 2008.

(2) On 14 December 2005, the London branch bought the right to the dividends from BNP Lux for £149,105,998. BNP Lux continued to own the shares.

(3) On 15 December 2005, the London branch sold the right to the dividends to an unconnected party, Alliance & Leicester Investments Limited (“ALIL”), for £150,000,000.

(4) At that time, the group also entered into funding and hedging arrangements with ALIL to enable HIL to meet its financial obligations and to hedge the group’s interest rate risk during the term of the on-going transaction (being the three-year period to the scheduled date for payment of the termination dividend). Under these arrangements: (a) HIL deposited £209 million of the funds it had received with BNP’s Dublin branch which agreed to pay interest on that amount over a three-year period at a fixed rate of 4.59%; and (b) the Dublin branch placed the £209 million with BNP’s Treasury function in return for floating rate of interest of 1 month Libor but agreed to swap that interest for interest calculated at a fixed rate of 4.74% on a total equivalent amount under swaps with ALIL’s parent, Alliance & Leicester plc (“A&L”), and BNP.

(5) BNP Lux exited from the transaction the following year. Under a put option entered into on 31 January 2006, BNP Lux sold the shares on 18 April 2006 to a UK subsidiary of BNP, BNP PUK Holdings Limited (“BNP UK”) for £62.7 million. BNP funded BNP UK by subscribing for shares in it for that amount.

(6) The arrangement with ALIL ended before the scheduled end date on 30 November 2007 when ALIL received the termination dividend of £150 million and the final fixed rate dividend. HMRC amendment

[5]Following a period of enquiry, HMRC issued a closure notice on 22 October 2010 in which they concluded that the London branch was not entitled to a deduction for the amount it paid BNP Lux to acquire the dividend rights. HMRC made an amendment by adding £149,106,000 (being the rounded-up amount of the purchase price paid for the dividend rights) back to the loss claimed in the amended corporation tax return. Initially, this was calculated incorrectly by reference to the wrong loss figure. HMRC corrected this following a review, as notified to the London branch in a letter of 25 February 2011, by adding the rounded-up purchase price to the correct figure of £96,091,000, to give a revised taxable profit of £53,015,000. The further facts relating to the issue of the closure notice are set out below. Summary of issues

[6]In brief, the first issue is whether the purchase and sale of the right to the dividends was made in the course of BNP’s banking and financial trade for corporation tax purposes. The parties take different views of the inferences to be drawn from the facts and the effect of the legal authorities.(1) In short, HMRC’s position was that, as a factual matter, the transaction was undertaken with the object of achieving the benefit of the tax loss by utilising the provisions of s 730(3) ICTA (we refer to this as the “s 730 benefit”). HMRC say that the authorities show that this means that the transaction was not undertaken by the London branch in the course of its trade and thus no tax deduction is available to the London branch for the price paid for the right to the dividends.

(2) BNP’s primary argument was that fiscal motive is not relevant in a case such as this: fiscal motive does not prevent what is, on an objective analysis, a trading transaction from being regarded as such. In any event, there are other commercial reasons for the transaction both from the perspective of the London branch alone and from that of the wider group. The London branch intended to and did make a significant profit from the purchase and sale of the dividend rights on a before-tax basis (of just over £400,000). In effect, the group borrowed £150 million at an attractive rate which resulted in a before-tax saving, after all costs, of just over £1.1 million and a considerable after-tax profit taking into account the s 730 benefit.

[7]Secondly, HMRC argued that, if it is held that the purchase and sale of the dividend rights was part of the London branch’s trade, the price paid by the London branch for the dividend rights was nevertheless not deductible for corporation tax purposes because it was not incurred wholly and exclusively for the purposes of the trade. BNP argued that a fiscal motive does not lead to this conclusion.

[8]Thirdly, the question is whether s 730(3) ICTA has the effect, as BNP argued, of excluding the sale price received by the London branch from ALIL from being brought into account as a taxable receipt of the London branch’s trade. In summary, under the provisions in place at the time, s 730(1) ICTA provided that, where the owner of shares sells the right to receive any distribution without selling the shares themselves, that distribution is treated as the income of the owner. However, under s 730(3):

“the proceeds of any subsequent sale or other realisation of the right to receive the distribution shall not, for any of the purposes of the Tax Acts, be regarded as the income of the seller or the person on whose behalf the right is otherwise realised.”

[9]It was common ground that the sale of the dividend rights by BNP Lux potentially falls within s 730(1) but the provision does not bite as BNP Lux is not within the scope of UK tax as a non-resident company which is not carrying on any trade in the UK. There was also no dispute that the sale of the dividend rights by the London branch to ALIL was a “subsequent sale” of the right to receive the dividend rights within the meaning of s 730(3). BNP argued that the meaning of the provision is entirely clear from the wording used; it simply excludes the proceeds from being brought into account for any tax purposes including as a receipt of a financial trader’s trade. HMRC argued that s 730(3) does not apply to a subsequent sale by a financial trader or that it only applies to prevent the proceeds of such a sale from being taxed as “pure income profit” in the trader’s hands. It does not, in their view, exclude receipts of a trade from being brought into account in the computation of the overall trading profit.

[10]Finally, BNP disputed that HMRC are entitled to raise the s 730 argument before the tribunal at all, on the basis that it was not within the scope of the conclusion or amendment made by the closure notice so that it is not part of the appeal. Evidence and facts

[11]We have found the facts set out below on the basis of the evidence given for the appellant by Mr James Peters and Mr François Demon, the evidence of Mr Marcus Stanton, a banking consultant, who produced an expert report at the request of HMRC, and the bundle of documents produced to the tribunal. Mr Peters is member of the Structured Capital Markets team of the London branch and Mr Demon is a member of the Global Equities and Commodity Derivatives section of BNP in Paris. The roles they had in relation to the transaction are further described below. They each produced witness statements which stood as their evidence in chief. Mr Peters was cross examined by Mr Goodfellow QC, who appeared with Mr Yates for HMRC, and Mr Demon was cross examined by Mr Yates. Mr Stanton was cross examined by Mr Boulton QC who appeared, as did Mr Flesch QC and Mr Jones, for BNP.

[12]The parties produced a statement of agreed facts which we have incorporated below. The parties dispute the interpretation to be given to and inferences to be drawn from the operation of the transaction and the relevant correspondence and documents evidencing internal discussions and processes at the time. Facts – the structure

[13]On 13 December 2005, HIL was incorporated as an investment company in the Cayman Islands and BNP Lux subscribed for 1.5 million ordinary shares of £1 each in HIL. BNP Lux paid a premium of £139 per share which gave a total subscription price of £210 million. Each share carried the right to:(1) a fixed rate dividend (payable monthly in advance) at the annual rate of 3.11% on the amount subscribed for the shares;

(2) a termination dividend of £100 per share payable out of the share premium account no later than 15 December 2008, which was £150 million in respect of all the shares; and

(3) participate in dividends declared at the discretion of the directors of HIL.

[14]The Memorandum of Association with which HIL was incorporated restricted its investment activities to:(1) depositing the sum of £209 million with the Dublin branch pursuant to the deposit agreement;

(2) investing up to £1 million in a portfolio of highly rated short dated sterling denominated debt securities;

(3) investing up to £7.5 million in sterling denominated senior ranking interest bearing deposits with BNP; and

(4) after payment of the termination dividend, investing in sterling denominated senior ranking interest bearing deposits with BNP or any of its affiliates.

[15]Article 11.2.7 of HIL’s Articles of Association contained provisions allowing any holder of at least 50% of the shares or of the rights to the dividends to require the termination dividend to be paid before the scheduled date on short notice of no less than three and no more than five business days:

“Any holder of at least 50% of the Ordinary Shares or, as the case may be, any Dividend Assignee with a holding of at least 50% of the rights to the Company’s Dividends, may at any time deliver an Acceleration Notice to the Company (with a copy to each of the Members and Dividend Assignees who did not sign that Acceleration Notice) signed by or on behalf of such holder or Dividend Assignee requiring the Company to pay the Termination Dividend on any date prior to the Scheduled Termination Date and being not less than 3 business Days nor more than 5 Business Days following the date of delivery of such Acceleration Notice.”

Consequently, both BNP Lux and ALIL (as the holder of the ordinary shares and as Dividend Assignee respectively) had the right to deliver an acceleration notice.

[16]HIL invested most of the funds received on the subscription by:(1) depositing £209 million with BNP’s Dublin branch on the basis that the branch agreed to pay a fixed rate of interest on the deposited funds of 4.59% over a three-year period; and

(2) purchasing £964,000 of Coca-Cola loan stock from the London branch. It appears from the correspondence that this stock was purchased so that HIL had UK assets to manage to ensure that it was regarded as UK tax resident.

[17]The Dublin branch placed the £209 million it received from BNP Lux on deposit with the Dublin section of the group’s Treasury unit on the basis that it would receive floating rate interest at 1 month Libor.

[18]The Dublin branch entered into interest rate swaps with A&L and BNP, which took effect from 15 December 2005, being the date when the external investor, ALIL, bought the dividend rights. Under the swaps, the Dublin branch agreed to pay floating rate interest at 1 month Libor to A&L, calculated on a notional sum of £150 million, and to BNP, calculated on a notional sum of £59 million, in each case in return for the receipt of fixed rate interest calculated on the relevant sum at 4.74%. The swap arrangements were put in place for the three-year scheduled term of the transaction with ALIL but could be terminated early with little or no breakage costs. Purchase and sale of dividend rights by London branch

[19]On 14 December 2005, under the terms of a written agreement entered into on that day, BNP Lux sold the dividend rights (but not the ordinary shares themselves) to the London branch for £149,105,998.

[20]On 15 December 2005, under a written agreement entered into on that day, the London branch sold the dividend rights to ALIL for £150,000,000.

[21]The agreement with ALIL provided for the London branch to give a number of representations and warranties. These included (under clause 6.1):(1) HIL (together with the directors of HIL) was in compliance in all respects with the Articles of Association.

(2) HIL had sterling denominated senior ranking interest bearing deposits with BNP of a maximum aggregate value of £7.5 million.

(3) The assets of HIL comprised only the above deposits and the deposit with BNP Dublin and other investments permitted under its Memorandum of Association.

(4) HIL had not incurred any material liabilities other than liabilities to pay tax on income on the above assets and other administrative expenses not exceeding £30,000 in any financial year of HIL.

(5) HIL was solely tax resident in the UK.

[22]Although not set out in the agreement, the London branch agreed to pay ALIL’s legal costs relating to the transaction. In effect, however, these were recouped by the BNP group through a small reduction in the fixed rate dividend. This was to ensure that ALIL was covered for the costs in the event of early termination of the arrangements (as is clear from the correspondence set out below). Sale of shares by BNP Lux

[23]On 31 January 2006:(1) BNP UK entered into a put option agreement with BNP Lux whereby it agreed that, on receipt of a put option notice, it would purchase the shares from BNP Lux for a total price of £62,700,000 (calculated at £41.80 per share). BNP Lux had the right to exercise the option by serving the notice between 6 April 2006 and 21 April 2006.

(2) BNP entered into an agreement with BNP Lux which provided that it would subscribe for shares in BNP UK, up to a maximum amount of £62,700,000, on being served a notice which BNP UK could serve at any time during the put option exercise period.

[24]BNP Lux served the put option notice on 13 April 2006 and BNP UK purchased the shares on 18 April 2006 for £62,700,000.

[25]In its accounts for the year ended 31 December 2006, BNP UK recorded a £2.7 million impairment against its equity investment in HIL and attributed to that equity investment a carrying value of £60 million as at 31 December 2006.

[26]On 30 November 2007, the arrangement with ALIL was terminated on the payment of a termination dividend of £150 million and a fixed rate dividend of £268,397.26 to ALIL on that date. Finance

[27]Mr Peters gave evidence that the BNP group parties were funded by the group’s Treasury unit, which was responsible for managing the group’s liquidity, through the dedicated Treasury team where the relevant entity/branch was located. The group’s business lines dealt with Treasury on an arm’s length basis as though borrowings from and deposits with Treasury were directly with the market. In 2005, there was a dedicated local Treasury team in each of London, Paris, Dublin and Luxembourg. The general rule was that any funding required or surplus liquidity generated by a business line within the BNP group should be drawn from or given to Treasury.

[28]Mr Peters went through BNP’s records as regards(a) transactions between BNP Lux and the Treasury unit in Luxembourg

(b) transactions between that unit and the Treasury units in other locations and

(c) transactions between the Dublin Treasury unit and the London Treasury unit in each case in the relevant period. From this he was confident that: (1) BNP Lux borrowed £210 million from the Luxemburg Treasury on 13 December 2005 which he could only assume was used to fund the subscription for the shares in HIL. (2) Luxembourg Treasury did not obtain the funds from London Treasury or any other Treasury function within BNP.

[29]He noted that it had also been confirmed to his colleagues in Luxembourg by the then head of Treasury in Luxembourg that in December 2005 the local liquidity position was largely positive and that it lent the majority of this net liquidity to the group. Based on net lending to the group, the outstanding position was in excess of €13 billion. Therefore, Luxembourg Treasury did not need to get funding from Treasury London, Dublin or Paris. The head also confirmed that, to the best of his knowledge, at the relevant time the Treasury in Luxembourg was “long” GBP (effectively it had a surplus of GBP). In any event, if it was short GBP then it would have been likely to enter into cross currency swaps rather than borrow the necessary GBP. Overall Mr Peters was confident that the £210 million used by BNP Lux to subscribe for the shares in HIL did not pass around the various Treasury functions in the BNP group in a circle. Group entities

[30]The London branch was, at all material times, a UK branch/permanent establishment of BNP which carried on its banking trade in the UK through the branch. BNP is a corporate entity which is tax resident in France and is subject to UK corporation tax only on the profits (wherever arising) attributable to the London branch as a permanent establishment in the UK through which it trades. The London branch carried on a trade of corporate and investment banking. In particular, in 2005 the London branch was the main centre of BNP group’s global capital markets business and of its bond issuance and secondary market trading of bonds. In December 2005, the two main business lines of the capital markets’ sector, the global Fixed Income and Global Equities and Commodity Derivatives business, had a head count of 720 people based at the London branch out of a group wide head count of 2,500.

[31]BNP Lux is a société anonyme incorporated in Luxembourg which is a subsidiary of BNP. It is mainly devoted to private banking, wealth management and related client services. It does not carry out significant capital markets activity. It is one of the BNP group’s two main private banking centres. This was the case in 2005 although these operations became much bigger following the acquisition of Fortis Bank by BNP in 2009. In 2005, it had around 515 employees, balance sheet assets of around €19.8 billion, turnover of approximately €142 million and post-tax profits of around €80 million. It was at all material times centrally managed and controlled outside the UK and was not resident in the UK for tax purposes.

[32]HIL was, at all material times, centrally managed and controlled in the UK by UK resident directors (who were employees of the BNP group) and was resident in the UK for tax purposes

[33]ALIL is a UK incorporated company which, at all relevant times, was UK tax resident and a member of the group of which A&L is the parent. Nature of the BNP group’s capital market activities

[34]The BNP group’s capital markets activities (including those carried out by the London branch) generally comprise activities in three categories:(1) buying, selling, lending and borrowing securities (such as bonds or equities) or rights attached to those securities,

(2) foreign exchange transactions, and

(3) derivatives contracts, where the underlying assets may be securities or baskets of securities (typically bonds or equities), indices, interest rates, commodity prices. One example of such derivatives transactions is dividend swaps, where one counterparty receives, and the other pays, an amount in cash equivalent to the dividends paid by a given company, against a fixed flow set at inception of the trade.

[35]Mr Demon gave evidence that, from a financial point of view, such transactions can be characterised as one or a series of cash flows. Such cash flows may be either determined or unknown at the time the transaction is entered into; they may be simple, standard transactions (often referred to as flow business) or more complex and bespoke (structured products). From the perspective of the clients and the counterparties of BNP, there can be a variety of drivers for entering into such transactions including managing their risks, raising financing or investing cash. As a party to the transactions (and not a mere agent), BNP is exposed to market risks arising from them, and must manage this exposure by entering into hedging transactions or promptly finding buyers for the instruments it has purchased. Both Mr Demon and Mr Peters expressed the view that the current transaction was a purchase and sale of cashflows such that it was made in the course of the banking trade of the London branch. We comment on this further below. Facts – overview of commercial and economic effects Purchase and sale of the dividend rights

[36]The London branch realised a profit of £893,373 (after taking into account its funding cost) on the purchase and sale of the dividend rights. The net figure, after deducting professional costs (taking into account recoverable VAT on the fees), was £402,372.

[37]As BNP Lux had invested £210 million in the shares but had sold all dividend rights for only just over £149 million, its investment in HIL was standing at a loss for accounting purposes as at 31 December 2005. That loss increased over the period it continued to own the shares in HIL due to the ongoing cost of funding the balance of its outstanding borrowing of around £59 million. As is clear from the evidence set out below, it was always envisaged that BNP Lux would sell the shares in HIL relatively soon after the dividend strip for a price which was set to ensure it recovered its funding costs and achieved a profit. It was for that reason that the put option arrangements were put in place in January 2006.

[38]The pricing of the sale of the dividend rights is discussed further below but essentially it reflects that the dividend rights comprised the right to receive £150 million after 3 years and, in the interim, to receive fixed rate dividends calculated at an effective rate of 4.354% on £150 million. This raises the question of why BNP Lux invested £210 million in the shares rather than £150 million. It is apparent from the economics of the transaction, as was confirmed by Mr Peters and is set out in the correspondence described below, that this was to ensure that HIL had sufficient funds to generate enough income to pay the fixed rate dividends on the shares and to fund its corporation tax liability on that income taking into account that the dividends were not tax deductible. On-going transaction with ALIL

[39]In effect, ALIL had made an investment in the BNP group of £150 million, which the appellant describes as the economic equivalent of a loan. ALIL received a return on this “loan” in the form of monthly dividend payments calculated at an effective fixed rate of 4.354% on a yearly basis. The principal of the “loan” of £150 million was payable as the termination dividend at the end of the three-year period. Mr Peters described this as equivalent to a bond carrying a fixed rate coupon. As the fixed rate dividends were paid between UK tax resident entities, they were not taxable in the hands of ALIL (under s 208 ICTA).

[40]HIL had received a total of £210 million of which £209 million was placed on deposit with the Dublin branch to generate interest income at a fixed rate of 4.59%. As noted, these receipts enabled it to have sufficient funds to pay the fixed rate dividends to ALIL and to pay corporation tax on the interest receipts. The fixed rate dividends were not tax deductible.

[41]The Dublin branch which was liable to pay fixed rate interest at 4.59% to HIL had a fully hedged position. It had in turn deposited the funds with the Treasury function in Dublin on the basis it would receive interest at 1 month Libor. Under the swaps, it exchanged that floating rate interest for fixed interest at 4.74% payable by A&L in respect of a notional balance of £150 million and from BNP in respect of a notional balance of £59 million. This meant that it had sufficient funds to pay interest to HIL at the lower rate of 4.34% and also realised a small surplus as regards the differential between the two rates.

[42]It is not disputed that the overall economic effect of these on-going arrangements with ALIL was that the BNP group received, in effect, a loan from ALIL at a rate which represented interest at 0.386% (or 38 basis points) below 1 month Libor. This low effective rate represents the difference between:(1) the rate of interest payable by A&L under the swap of 4.74% on a notional loan of £150 million resulting in payments due to the BNP group of £7,110,000 for each year; and

(2) the effective dividend rate of 4.354% due to ALIL from HIL on the “loan” of £150 million made by ALIL resulting in payments due from the group of £6,531,000 for each year.

[43]The envisaged total pre-tax saving for the BNP group, as a result of this differential in interest rates was £579,000 per year, being the difference between £7,110,000 and £6,531,000. If the transaction had run its course over three years this would have given a total pre-tax saving of £1,737,000. As the arrangement with ALIL was terminated on 30 November 2007 (after 715 days) the actual pre-tax profit for the BNP group was £1,134,205 (before deducting costs relating to the transaction).

[44]Mr Stanton set out in his report that the headline rate of funding and saving achieved did not take into account the lack of a tax deduction for HIL for the dividend payments. If it did, the coupon could be expected to be grossed up to a pre-tax equivalent of 6.21%. From this perspective, the headline borrowing rate, absent the s 730 benefit, would be Libor plus 1.47%. This would be well above the BNP group’s normal cost of borrowing which Mr Stanton thought would be in the range of Libor plus 0.1% based on the rounded credit default swap rate for BNP taken from Bloomberg’s database as at December 2005. It was not disputed, therefore, that the transaction would have been loss making on an after-tax basis but for the obtaining of the s 730 benefit. The loss-making nature of the transaction, absent the s730 benefit, is acknowledged in the correspondence and evidence set out below.

[45]Mr Peters provided a calculation of the expected pre-tax return from the transaction for each entity and the actual return which was compiled from the bank’s records. It was not possible to verify independently all the underlying cash flows so that some extrapolation was necessary to reconcile the expected return with the actual. The figures were reconciled within £36,141 which in a transaction of this size was considered immaterial. The figures were not disputed by Mr Stanton. Expected return Actual return BNP (8,168,271) (7,323.369) Dublin branch 561,552 561,551 London branch 912,476 893,373 BNP UK 0 0 BNP Lux 1,814,471 847,248 HIL 6,013,978 6,118,978 1,134,205 1,097,781 Sale of shares in HIL by BNP Lux

[46]The shares were sold by BNP Lux to BNP UK at an overvalue for a sum of £62.7 million. As noted, this was to ensure that BNP Lux recovered its funding costs and made an overall profit. The profit was £750,000 after costs (and before tax).

[47]BNP UK recognised an impairment loss due to the overpayment of £2.7 million. As BNP financed the payment for the shares in HIL through subscribing for shares in BNP UK, it had an on-going funding cost on £62.7 million. Mr Peters said that £59 million of this amount accrued interest at a floating rate and that interest was effectively swapped into fixed rate interest under the swap with BNP Dublin. He thought that the remaining balance of £3.7 million would also have accrued interest at floating rate but he had not been able to find a specific booking to confirm this or to confirm the amount of interest. Disputed areas

[48]There is no material dispute on the actual economic effects of the transaction as set out above. The central factual issue is whether the London branch effected the dividend strip in the course of its (undisputed) substantial banking trade. The dispute concerns the inferences to be drawn from and characterisation of certain aspects of the transaction in view of its economic effects and the resulting impact on this central factual issue.

[49]In particular, there was much debate about(a) the characterisation of the sale of the dividend strip by BNP Lux at a price which HMRC argued was inherently loss making for it and the subsequent sale of the “rump” at an overvalue and

(b) whether the “loan” from ALIL could really be said to be made at a cheap rate from the BNP group’s perspective. These issues are considered below. Facts – implementation of the transaction Origin of transaction

[50]The Optimisation Finance (“OF”) team in London was the unit within BNP which put together the transaction. The key members of the team were Mr Peters, Mr Nick Williams and Mr Stuart Bayfield. The head of the team in London was Mr Neil Robinson who reported to the group head based in Paris, Mr Delafontaine.

[51]Mr Peters was involved throughout the process but his responsibility was primarily for the technical analysis, structuring and approvals process and his involvement diminished once the transaction was in the negotiation and execution phase. Mr Williams drove the negotiation with ALIL and Mr Bayfield drove the operational set up, trade bookings and management. Mr Peters has worked at BNP since 1990 initially on secondment from his then employer, KPMG, and on a full time basis since 2001. He joined OF on 1 January 2005 having previously worked in the tax group. Mr Williams and Mr Bayfield have subsequently left BNP. The transaction was given the project name “Sumatra”.

[52]The idea first came to the attention of the OF team in London through presentations made to them by two other financial institutions (Swiss Re and Dresdner) in the middle of October 2005. The proposals shared the same essential features as that adopted by the London branch; the key elements were the creation of a dividend right initially held by an offshore entity, the sale of that right to the London branch and the onward sale of that right to an external investor. Mr Peters confirmed that this was where the idea came from for a transaction that used s 730 to generate a tax benefit. He could not remember specifically what had happened but generally when presented with such a proposal the team would have analysed it to see whether it was commercially viable. The team would not recommend or push forward a transaction which had a significant tax effect unless there was a commercial element. From looking at the presentation papers produced by Swiss Re and Dresdner, Mr Peters could not see how those proposals would stand up from a commercial perspective. He said that: “the transaction that we ultimately presented is different, in that there was a strong commercial rationale behind that and that was the fund raising from the ALIL group”.

[53]Therefore, after the idea for obtaining a tax benefit through the operation of s 730 came to their attention, the OF team decided to work up the proposal into a form which, as Mr Peters said, they considered “commercially sensible that may be able to take advantage of this [the s 730 benefit]”. It is clear from the origination of the proposal and what happened subsequently that the team set about putting together a structure which had what they considered were the necessary commercial elements for the structure to work to obtain the intended tax benefit. As Mr Peters said repeatedly in his oral evidence, each party to the arrangements had to be “commercially incentivised” as otherwise the tax planning would not work. This was not a case of the bank seeking funding which it then structured to optimise the tax effect. Rather the team took a tax planning idea and built around it to introduce “commercial” elements to give it what they considered to be the best chance of succeeding.

[54]That obtaining a substantial tax benefit was the key motivator is apparent from the early stages and throughout the internal correspondence. For example, on 26 October 2005, the head of Optimisation Finance, Mr Robinson emailed a colleague attaching a structure paper for the proposal noting that:

“it is designed to create a one-off deduction shortly after implementation.”

[55]He continued to identify that as the “critical issue” before deciding to pursue the transaction was finding a tax investor to whom the dividend strip could be sold as the London branch would need to make a profit for the planning to work and this would provide the commercial rationale for that profit:

“The reason we are targeting tax investors is to ensure that we will make a profit in London branch (a tax requirement) when buying the dividend strip (probably from a group entity in Ireland) and selling it to the UK investor. As the strip will be received tax free in the hands of the investor they will be prepared to pay more for it than the London branch will have to pay when acquiring it.”

[56]That this was the reason for targeting UK tax paying investors accords with Mr Peters’ evidence. Mr Peters accepted that it was important for the London branch to make a profit from the purchase and sale of the dividend strip for the tax planning to work. He said that was important to the tax analysis as it “had to make sense to all entities, all parts of the group”, to enter into the transaction. He agreed that otherwise the transaction would have been more vulnerable to challenge on the basis that it was not a commercial trading transaction. Therefore, the pricing was set to give London branch a profit from the purchase and sale; essentially the sale price to ALIL was fixed and the purchase price had to be lower. In his witness statement, he gave the differential in the tax position of the London branch and ALIL as regards the dividends as a justification for the difference in the price paid by the two parties for the right to the dividend.

[57]Mr Robinson concluded his email by noting that:

“it is a bit of a long shot but if we can find an investor we may still be in a position to generate a significant tax deduction in 2005....we would ideally wish to see the details of the Autumn Statement (expected sometime in November) and to source an investor before seeking to implement.”

[58]The desire for the transaction to be done quickly and before the end of 2005 is evident in numerous places throughout the correspondence. It is apparent that this was due to the desire to generate the tax benefit in that accounting period to use the tax capacity of the group for that year and because of a concern that there was a high risk of a change to the law given that the planning utilised what was accepted was an unintended effect of s

[730]59. It seems that by late October 2005, the OF team had worked up the proposal sufficiently that it was discussed with the London tax team. On 31 October 2005, an outline of the transaction was circulated more widely within the BNP group by Mr Bayfield who noted that this was a transaction which the team were “keen to progress quickly, and we are currently in the process of sourcing an investor.”

[60]The outline transaction document is a draft of the document prepared by OF which, when finalised, was submitted (together with documents prepared by the other relevant departments within the BNP group) to the relevant committees within the group whose approval was required for the transaction to go ahead. In this document and related correspondence, the transaction is presented in headline terms as a low cost funding opportunity. The initial draft is headed - “An opportunity for the [BNP] group to raise low cost funds from selling a dividend strip.” In the executive summary under the heading “Commercial rationale for the transaction” it is stated that “[OF] have identified a transaction which will allow the [BNP group] to raise £100 million at an attractive post-tax cost of funds”.

[61]On the same day, in response to a request from Mr Richard Burge for further information on what transactions were in the pipeline, Mr Bayfield responded:

“This is very much “hot off the press”. This is a transaction which we have come across in the last couple of weeks, and has the best fit for [BNP] if we can close it out before year end. Having discussed it with Group Tax we have sent out the attached email with the intention of getting all of the functions up to speed as quickly as possible.”

Finding a UK investor

[62]By late October 2005, the team were contacting potential UK investors. Mr Peters stated that the OF team would generally contact clients with the assistance of the relevant members of the London client relationship team. In some cases, where the OF team had existing relationships with a client, they would contact the client themselves. Marketing materials were prepared for 19 potential counterparties although Mr Peters was not sure if they were all in fact sent out. He was also aware that three other parties were contacted but, as they expressed no interest, marketing materials were not prepared for them. Mr Peters confirmed that, as indicated in the email sent by Mr Robinson, the team targeted UK tax paying companies which would potentially see the opportunity to invest in return for a tax free dividend return as attractive.

[63]The email correspondence in late October/early November 2005 indicates that the team felt they were not having much success in finding an investor. Approaches had been made unsuccessfully to United Utilities and International Power (and at least one other unnamed potential counterparty) and consideration was being given to whether to seek to effect the transaction under reciprocal arrangements with Fortis and/or Dresdner banks. As regards Dresdner, a concern was raised by Mr Bayfield that they may think “wrongly, that we stole the technology from them” and that when they had proposed this they were looking for a “tax capacity investor (which we cannot do) and not a credit investor”. The response from Mr Robinson was that “Oke gave them the impression we had the technology. We should be ok.”

[64]However, in that correspondence, on 1 November 2005 it was reported that A&L had expressed an interest in the transaction and from around this time onwards the team focussed their efforts on A&L. The contact was made direct by Mr Williams who had an existing good relationship with contacts at A&L. He initially contacted Rachel Morrison who worked in A&L’s tax department. A presentation document had by then been sent to them and, in an email of that date, Mr Williams commented “This appears to be just right for them. They will revert on Monday with a yes/no. I have a good feeling. They want to also close this year.” Initial proposal using BNP Ireland

[65]At this stage, it was envisaged that the offshore company, which would subscribe for shares in the new company, would be a subsidiary of BNP in Ireland, that it would invest around £140 million in the new company, that it would sell the dividend rights to the London branch for £99,500,000 and that the London branch would sell the rights to an unconnected UK investor for £100,000,000. In an email of 3 November 2005, Mr Bayfield set out the potential benefits of the transaction for the UK tax investor and the BNP group as follows:

“The investor will receive a tax free stream of dividends from the issue which will carry a PTE margin of circa 143 basis points. Over 2 years this will equate to a pre-tax equivalent yield of 3m. The [BNP] group will have a one-off tax deduction at the commencement of the transaction of circa 100m. If we assume that of this £100 million deduction, £8.3 is used to shelter the tax capacity created by paying non-deductible dividends to UK investor, this leaves a net deduction of 91.7m available to the group. This equates to a tax saving of 27.4m, which on a pre-tax equivalent basis equals circa 39m.”

[66]Mr Peters confirmed that this earlier proposal had essentially the same features as that which was implemented (albeit that the transaction size was smaller). He confirmed that the reason that the proposed investment by the offshore company in the new company was £140 million rather than £100 million was to ensure that the new company would have sufficient funds to pay the non-tax deductible fixed rate dividends and tax costs. This was why an equivalent proportionate amount was added to the investment in the actual transaction so that rather than investing £150 million BNP Lux invested £210 million.

[67]It was pointed out to Mr Peters that right from these early stages it was envisaged that the offshore company (then envisaged as BNP Ireland) would sell the “rump” of its interest, being the shares in the new company once the dividend rights had been sold, to a UK member of the group. The paper setting out the original proposal contained the following statement:

“During 2006, it is anticipated that Ireland will sell the ords to a UK tax resident subsidiary of the group for £500,000 profit, and that UK InvestCo will fund the investment from new equity from France.”

[68]Mr Peters confirmed that was always part of the proposal as the offshore seller of the dividend rights had to be “incentivised”:

“the dividend vendor had to be incentivised to make, you know, a commercial return from his involvement in the wider transaction and I think quite early on it was clear that would be through the sale of the rump.”

[69]In a later email of 9 November 2005, Mr Williams reported that BNP had two UK investors seeking to invest £100m each, one for 2005 and one for

[2006]He noted that for the transaction to work, new equity would be needed from BNP but that he assumed that would not be a problem given the potential tax benefits of the transaction for the group:

“… we will require an injection of new equity from France into the BNP UK group. Per GBP 100 mln of transaction we will require approximately GBP 40mln of new equity. The reason for new equity is a UK company law corporate governance issue which I can explain in more detail. As the transaction is so remunerative for the [BNP] group (overall expected after tax profitability per GBP 100mln being around GBP 29mln), I assume that further capitalising the [BNP] UK group to the extent required will not be an issue. The commercial rationale being the raising of attractive term funding for the Bank. If that is not the case, please let me know as soon as you can (before we start to incur significant costs of due diligence).”

[70]Mr Peters again agreed that this demonstrated that the intention was that the “rump” of the shares would be sold by the offshore company and that the team “were going to try to deliver that for [BNP Lux]” but noted that “there was no guarantee upfront”. Involvement of UK tax and commencement of approvals process

[71]By the middle of November 2005, the BNP internal tax team had become involved in looking at the proposal. On 11 November 2005, the head of UK tax, Mr Peter Scholes, emailed Mr Demon with his initial views on the transaction:

“Optimisation Finance have timetabled this to close before the end of the year. My initial view is that the proposal is technically very strong, and it may turn out to be one of the last principal deduction opportunities we have in the UK. Apparently [A&L] are lined up to do this before the end of the year (subject to approvals etc). I expect the fact that they have a UK listing to be helpful to the UK tax analysis. OF also have a potential client for a second deal in the early new year (possibly with in excess of £300m to invest). One important factor is the size. OF’s original proposal for the A&L deal was for an amount of approximately £99m (consideration paid by London Branch for the dividend strip; giving rise to an [sic] one-off UK tax deduction, reducing 2005 capacity). However, A&L may be keen to invest more – and it [sic] I think we should decide what size of deal we would contemplate for both 2005 and 2006. NB Michael’s 2005 tax capacity figures currently show £150m taxable profits in the branch with a further net £100m in London … £36m outside London and £68m of one-off deductions (including 100% of Zephyr). This nets to £218m before Sumatra I. The current global estimate for 2006 is about £300m (transactions already under discussion or already implemented could utilise between £90m and £150m of this (before Sumatra II)).”

[72]Mr Scholes here indicates that the size of the transaction may be determined by the tax capacity of the group in the relevant period and that this was in fact the case is clear from the later correspondence. Mr Peters agreed that there was no point to the transaction unless the London branch had tax capacity to shelter as otherwise there was no real tax benefit.

[73]It was put to Mr Demon that Mr Scholes clearly envisaged in this email that there would be another transaction of the same type and that the reason that OF did not pursue such further negotiations was the later change in law which prevented the tax planning working. Mr Demon did not accept this although he gave no convincing reason to the contrary. He said that it was a long way from the above email to another potential Sumatra transaction and he did not think “we can jump to” that conclusion given another transaction was only a possibility at that stage. Our view is that it is not credible that the prospect of pursuing the possibility of entering into another transaction of the same type was dropped for any reason other than the change in law given the repeated references in the correspondence to the need to effect the transaction quickly and the concern that there may be something in the pre-budget report which would affect the potential to obtain the tax benefit.

[74]By 14 November 2005, the relevant teams had started to produce drafts of papers which would need to be presented to the various committees within the BNP group whose approval would be needed for the transaction to go ahead. The approvals process is set out in detail below.

[75]By 16 November 2005, it had become apparent that the transaction would not proceed with BNP Ireland as the offshore company in the structure. On that day, Mr Peters sent an email stating that there was an Irish tax problem in using BNP Ireland “which could not be resolved without putting a great deal of stress on other aspects of the transaction. We have therefore decided to seek another jurisdiction (almost certainly Lux or Neths)”. On the following day, Mr Peters made the first contact with BNP Lux as set out below.

[76]On the same day, Nick Williams made an enquiry as to who he could ask in BNP’s middle office team about the booking of the transaction from the London branch’s perspective. He stated that he wanted to book it in a fixed income front office system in the same way as any other fixed income security as this “enhances our facts”:

“As a trading transaction for [BNP] London branch, I want to book the acquisition and sale of the dividend strip in a Fixed Income front office trading system (in an [Optimisation Finance segregated book]) in the same way as any other fixed income security. The bond trading system BND is the preferred option. Our ability to book the deal in this way enhances our facts.”

[77]This continued to be a concern, as on 22 November 2005, he emailed Mr Paul Smith stating that:

“We need to be able to book the purchase and subsequent sale [of the dividends] in an [Optimisation Finance] dedicated book where we can clearly show records that support the fact that the acquisition cost was £99.5m and the sales proceeds were £100m.”

[78]On 17 November 2005, there was a meeting between BNP and A&L and their respective lawyers. In an email following that meeting, Mr Robinson reported to Mr Delafontaine that the meeting had taken place and that:

“the biggest obstacle is the Pre-Budget statement (found out today that this is on 5th December). Fingers crossed!”

[79]In discussing this correspondence, Mr Peters said that he suspected that, as Mr Robinson seems to suggest in this email, the transaction would not have gone ahead if there had been a change in the law in the pre-budget statement affecting the availability of the s730 benefit. He thought that the transaction would not have gone ahead unless; “there was a deduction – unless the wider transaction achieved the deduction at least equal to the cost of the dividends that were paid to the [A&L] group, then, on a post-tax basis, it would not have made commercial sense to us. That is what we needed.”

[80]Mr Peters sent an email to Mr Scholes on that day noting that 19 clients had been approached regarding the proposal of which four were still interested and one was able to close before the year end. He said that it was hoped to transact with the others in early 2006. Involvement of BNP Lux

[81]From 17 November 2005 onwards, there was much correspondence between the OF team and various individuals at BNP Lux regarding the involvement of BNP Lux in the transaction. It appears that the first contact was made by Mr Peters who contacted Mr Eric Berg of the Luxembourg tax team on 17 November 2015 by email attaching the OF outline approval document. In the email, he noted the difficulty with Ireland and stated that OF had been advised that Luxembourg should work and they were very keen to do the deal with BNP Lux.

[82]The principal personnel at BNP Lux who were involved in the discussions on the proposal were Mr Berg, Ms Stephanie Majchrzak Gilot (who headed the structured transactions team in Luxembourg) and Mr Yvan Juchem (who was responsible for finance). Mr Peters described Mr Berg and Mr Juchem as having a role within “legal entity management” which was not a client-facing role. In his witness statement Mr Peters said that, so far as he knew, none of the personnel at BNP Lux were involved in the development or marketing of the transaction or the negotiations with ALIL and Ms Majchrzak Gilot had confirmed to him that she also believed that to be the case. The employees of London branch were responsible for the development and marketing and had carried out the negotiations with ALIL (primarily through Mr Williams) and BNP Lux in parallel. Mr Demon also gave evidence that it was clear to him in his role as head of GF that the London branch lead the deal: it developed the structure, marketed it to counterparties, found ALIL, negotiated with ALIL and lead the implementation process.

[83]We accept that BNP Lux were not involved directly with ALIL. The views expressed by Mr Peters and Mr Demon are supported by the correspondence from which it is clear that the proposal originated from the OF team in London who were the driving force behind the transaction. As set out in further detail below, it is clear that the team at BNP Lux took a very active interest in what was proposed, that they wanted to safeguard its position and they were keen to earn a profit from its involvement. It is equally clear, however, that they entered into the transaction at the request of the London branch with a view to securing the desired tax benefit for the group.

[84]Mr Peters said in his witness statement that, as the team at BNP Lux did not have the market profile or the history of putting together such complex arrangements, he thought it was never contemplated that ALIL would purchase the dividend rights direct from that entity. In his experience, counterparties prefer to transact with the entities in the group with the highest creditworthiness, which in this case was BNP. The sale and purchase agreement with ALIL contained a number of representations and warranties given by the London branch that HIL would run its affairs in such a way that it would be able to pay the dividends. He worked on the basis that these warranties and representations were critical to ALIL.

[85]Mr Demon similarly stated in his witness evidence that the idea that BNP Lux could have sold the dividend rights direct to ALIL was unrealistic. BNP Lux did not have a capital markets business so he would not expect it to have the appetite or the expertise to put this type of structure in place or access to the type of clients who would want to enter into it. Moreover, the London branch had an existing relationship with A&L which made it unlikely that BNP Lux could have transacted with them without the involvement of the London branch. Even if BNP Lux could have transacted direct, it could not have secured the same terms. In the event of default, ALIL would have had recourse only to that entity’s assets which are not as substantial as those of BNP, it did not have its own credit rating and in his experience counterparties prefer to transact with the parent company from a credit risk perspective and, for similar reasons, with a credit-rated entity. If BNP Lux had sold the dividend rights direct, he would expect ALIL to have agreed to pay only a lower price or to require BNP to give a guarantee for which BNP Lux would have had to pay a fee.

[86]On the credit point, Mr Stanton explained in his report that, as a general matter, the price which a third party would be expected to pay for an income stream would reflect the creditworthiness of the payer. In this case, ALIL was able, in effect, to look to the credit strength of BNP as HIL’s main asset (the funds of £209 million) was placed on deposit with it (at its Dublin branch) and the position was supported by the representations and warranties given by the London branch. He noted that it was reasonable to suppose that BNP Lux would have had a lower credit rating than BNP on a stand-alone basis so justifying a difference in price on a sale of the right to the dividend by BNP Lux. However, he noted that if BNP had provided parental support then the credit rating could be expected to be the same and, accordingly, it could be expected that it would achieve the same price if it had sold direct to ALIL as the London branch obtained.

[87]Mr Peters made some general comments in his second witness statement on BNP’s approach to assessing the creditworthiness of entities it proposes to do business with. The overall tenor of this was that a subsidiary would not automatically be regarded as having the same credit rating as its parent. It may be upgraded where there was a strong reason for presumption of parental support but ordinarily only a written guarantee would suffice for the same rating to apply. He also noted that other variables can be used in conjunction with the credit rating to assess risk; one such variable being “risk-adjusted return on capital”. He noted that, under this test, a parent could clear the required hurdle whereas a subsidiary may not. Although it might be expected that a parent would stand behind its subsidiary, this was not a given and BNP’s approach recognises this by assessing the credit rating for each entity separately and recognising that the risks associated with dealing with a subsidiary are different from those associated with dealing with a parent. The internal credit rating assigned to clients of BNP would therefore generally reflect the increased risk of dealing with subsidiaries (whether guaranteed or otherwise) as opposed to their parent companies.

[88]Mr Peters accepted, in his oral evidence, that the group would not have achieved the s 730 benefit if BNP Lux sold the dividend rights directly to ALIL. He agreed that if matters had been structured differently, such as by the provision of a parental guarantee by BNP, BNP Lux (or any entity in the BNP group) could have sold the strip to ALIL. Given that the major asset of HIL was a deposit with the parent company, if that deposit was in danger, then the personal covenant of the bank was in danger as well. He said that if BNP was in bankruptcy then he doubted that the covenants would be worth very much more than the recovery of the loan. Mr Stanton gave a similar view as regards the warranties.

[89]We accept that, given the very different roles of the London branch and BNP Lux and their differing areas of expertise, it is highly unlikely that BNP Lux would want to or be able to put together this type of transaction itself. We do not accept, however, that there is any particular reason why the London branch could not in principle have arranged a sale by BNP Lux direct to a third party such as ALIL on the same or similar terms as those achieved by the London branch given the views of Mr Stanton and Mr Peters on the credit position.

[90]In any event, it is entirely clear that there was no intention for BNP Lux to sell the dividend strip direct to ALIL. As Mr Peters accepted, there was no reason for the inclusion of BNP Lux in the structure, as the initial seller of the dividend strip, except to be able to take advantage of the perceived unintended effect of s 730. London branch could not have obtained the benefit under this structure if it itself had invested in the shares in HIL and sold them. There needed to be a prior sale to it and that needed to be made by an offshore entity located in a favourable tax jurisdiction so that it did not suffer any UK tax charge on the sale. Discussions with BNP Lux

[91]On 18 November 2005, Mr Peters emailed Ms Majchrzak Gilot of BNP Lux setting out the proposal in more detail, namely, at this stage that BNP Lux would acquire £140 million of shares in a newco for £140 million, BNP Lux would then sell the right to the dividends due over two years to BNP London which would pay £99.5 million for those rights. He went on to consider the position as regards the “rump”, meaning BNP Lux’s holding of the shares once denuded of the dividend rights, following the sale of those rights:

“It is expected that the “rump” of the shares after the dividend rights have been sold will have a fair value of approximately £40 million. [BNP Lux] would use the proceeds to repay most if its original borrowing taken out to fund the subscription. At the end of 2005 we would expect [BNP Lux] to have a liability (the intra group debt) of £40.5 m plus any accrued interest [ie £140m - £99.5m]. The company would also have an asset (the shares) which we believe would have a fair value of £40m. The difference between the liability of £40.5m and the asset of £40m should be an accounting loss of £0.5m. To put it another way, [BNP Lux] would have paid £140m for the asset and disposed of a part of it for £99.5m. The remaining asset would be worth £40m and therefore a loss of £0.5m should arise [ie £140m costs - £99.5m sale proceeds - £40m residual value = £0.5 m loss]. During January 2005 [BNP Lux] would sell the remaining “rump” of the shares to a [BNP] UK company for approximately £41m. The sale price would be calculated to ensure that, after all costs, a profit will arise in Lux as a result of these arrangements. Although we have yet to determine exactly how much profit should be left in [BNP Lux] (£500,000 may be too much) the principle remains that [BNP Lux] will make a commercial profit overall from its involvement in the arrangements.”

[92]In his oral evidence, Mr Peters agreed that unless BNP Lux could sell both the dividend rights and the “rump”, being the shares, the transaction would be disadvantageous to it. It was never the intention that BNP Lux would hold the shares in the long term and the idea that was sold to BNP Lux was “a composite package; it would sell the strip and ultimately would sell the rump”. Therefore, the intention was that the “rump” would be sold for a value that would ensure that BNP Lux made a profit from its involvement, essentially, at an overvalue.

[93]He agreed that BNP Lux was “fully expecting to be bought out at the end” but said that “it was not guaranteed upfront. It was clearly contemplated and we within [OF] would have done everything we could to ensure that [BNP Lux] was bought out”.

[94]He described this as “not a racing certainty” but said that “we would be very comfortable and Luxembourg would be very comfortable that we would able to deliver that, hence they would not have entered into it … so in practical terms, it was highly likely”.

[95]He said that although BNP UK was potentially a target (to be the buyer of the rump), it was never guaranteed to BNP Lux that would be the case. The board of directors of BNP UK were not contacted upfront although it was intended that would happen. He agreed that the reason for the delay was for tax purposes as is apparent from the later correspondence discussed below.

[96]When questioned about whether the price for the sale of the dividend rights was a commercial one from BNP Lux’s perspective, given that it was essentially in a loss making position at the point of sale, he said:

“I don’t believe it’s a case of Luxembourg not selling the strip were it not be able to sell the rump. I think Luxembourg would not have entered into the transaction at all and capitalised [HIL].”

[97]Mr Peters also agreed that, as indicated by his email, the pricing was to be arranged so that both London branch and BNP Lux made a profit albeit that the precise level of profit BNP Lux would get was probably subject to negotiation.

[98]On 21 November 2005:(1) Mr Peters sent Mr Berg a request for BNP Lux to obtain an opinion from Clifford Chance on the Luxembourg tax position. He said that: “where possible, we would prefer [BNP Lux] to pay for all advice directly attributable to the company. The transaction proposed will be structured so as to ensure that after all costs (legal, tax, accounting, funding etc) [BNP Lux] will have made a profit. Although we have not determined precisely how much, I would expect this profit to be in the region of GBP 200k. We will, however, agree what level of profit is acceptable to Luxembourg with you before closing”.

(2) Ms Majchrzak Gilot (who was copied in to this email) responded that she expected BNP Lux’s profit to be £500,000 as had been stated initially and asked for this point to be confirmed before Mr Peters went ahead.

[99]It was put to Mr Peters that he had made the point about the costs as he wanted to preserve the margin that the London branch would earn from the transaction. He said that he thought the costs in question were costs attributable to BNP Lux which that entity should pay in the normal course but he agreed that he “wanted as few costs in London as possible”.

[100]Over the next couple of days there was correspondence between Mr Berg, Mr Peters and Mr Scholes regarding the request for an opinion from Clifford Chance in Luxembourg. Mr Berg was concerned that the usual procedure was not being followed in that it appeared Clifford Chance were being instructed without the obtaining of a fee quote or alternative quotes and without the involvement of group tax.

[101]Mr Scholes said that, given that Clifford Chance were already involved in the UK and the very short timetable for analysing and implementing the transaction, he was happy for Clifford Chance in Luxembourg to provide the advice as long as Mr Berg had no objection. He said:

“Whilst I appreciate there is normally a requirement to put advice out to tender given the circumstances of this transaction (this is a very big opportunity for the [BNP] group with limited implementation time), I would like to authorise optimisation finance to appoint Clifford Chance to act…. Subject to receiving a reasonable fee quote, I am very keen for this to proceed as quickly as possible. The potential UK tax benefit is potentially some million if the transaction can be approved and implemented before the year end.”

[102]It was put to Mr Peters that this demonstrates again the desire for this to be done before the end of the year and before a change in law so that the tax benefit could be obtained. Mr Peters essentially agreed that was the case.

[103]It seems the fee quote was obtained eventually by Ms Majchrzak Gilot. In an email of 23 November 2016, Mr Peters thanked Ms Majchrzak Gilot for having obtained the fee quote and stated that he was happy for her to go ahead and instruct Clifford Chance on behalf of OF on the basis of the quote.

[104]On 21 November 2005, Mr Peters also responded to a query from Mr Daniel Sladen in the UK tax team on the level of profit to be made in each BNP entity involved in the transaction, including BNP Lux. Again, this shows that the transaction was structured to give each party a commercial “incentive” for entering into the transaction:

“[BNP Lux] – Acquires shares for 140m. Sells strip and corpus for 99.5m and 41m respectively. Should make a profit of 0.5m. [London branch] – Acquires Dividend Rights for £99.5m and sells for 100m making a profit of 0.5m. [BNP UK] – Acquires rump for 41m. Although there is no return expected for 2 years, the acquisition is funded out of new equity and therefore the company should be seen to be sufficiently incentivised (from a corporate governance and tax point of view to enter the transaction). [BNP] – In consolidated accounts shows pre-tax profit of 35bps per annum. This is the difference between the dividends paid (shown as interest payable) at 4.15% on 100m and the interest received on the 100m (assumed to be 4.5%). Each party is therefore incentivised (absent tax) on a pre-tax basis to enter into the arrangements.”

Correspondence in late November 2005

[105]It is apparent from the correspondence that, by late November 2005, matters were progressing with A&L.

[106]On 19 November 2005, Mr Bayfield sent an email to Mr Williams stating that the most rational thing for BNP would be to exit the transaction almost immediately and wondering if A&L realised this. He said:

“I know I keep going on about this point, however, it would seem that the most rational thing for the [BNP group] to do is to pull the transaction on day 2 because we will have secured our tax advantage, and therefore paying [dividends] to [A&L] at the rate agreed is a drag on the group. Obviously, there is no current intention to do that given our relationship with [A&L]. However, do you think that it is completely clear in [A&L’s] mind we are losing money after day 2, but that the only reason…we will be staying in the transaction is the strength of our relationship.”

[107]On 22 November 2005, Mr Williams sent an email to A&L with a timetable for agreeing documents with them and noting that he would send a second draft of the term sheet to them later that day following the consultation with tax counsel, Mr Flesch. He said it would be helpful if A&L and their advisers could have a quick look and identify any “bits you do not like”.

[108]In correspondence of 23 November 2005:(1) Nick Williams reported that A&L wanted to do a three-year deal.

(2) Mr Scholes sent Astrid Hosxe of the group tax team a note with the UK tax analysis to assist with that team’s review of the proposal. He said that “overall, my view is that the UK tax analysis is strong enough for us to do the transaction, especially in view of the huge upside and relatively small downside.”

(3) Mr Peters emailed Ms Majchrzak Gilot regarding the fee quote as set out above.

[109]In late November 2005, it was becoming apparent that there was a tension between the desire on the part of BNP Lux to exit from the transaction at an early stage and, as was confirmed by Mr Peters and the later correspondence, the advice of the tax team that the exit should be delayed until the next fiscal year. It seems (from the correspondence and Mr Peters’ evidence) that the desire in Luxembourg for this to become definite in January was at least in part due to a regulatory issue that certain approvals would be needed from the regulator in Luxembourg if the shares in BNP Lux were held for more than three months.

[110]On 24 November, Ms Majchrzak Gilot asked Mr Peters if he could confirm that BNP Lux would exit from the transaction in January 2006. He responded that:

“it would be difficult to provide a definite offer by a UK company to acquire the shares in [HIL] on a specific date in 2006. This is because it would put some tension on the analysis of the transaction. Although we will seek all internal approvals on the basis of the shares being sold to the UK in early 2006, we do not want to approach the directors of the UK company in question or to determine transfer dates or actual sales proceeds until after the Dividend Rights have been sold to UK Investor.”

[111]Mr Peters accepted that the tension on the transaction which he referred to as preventing the sale of the “rump” being definitely put in place in advance was in relation to UK tax although he could not remember the specifics of the tax issues.

[112]Ms Majchrzak Gilot also queried (it seems on a call with Mr Peters as referred to in his email) why it was necessary for BNP Lux to invest £140 million in the transaction given that the dividend rights were only in respect of £100 million. Mr Peters responded in the same email that BNP Lux needed to invest £140 million in the transaction so that the new company would have sufficient distributable reserves to pay the return on the investment. He went on to explain the pricing of the different parts of the transaction as follows:

“The price that the London branch is prepared to pay is basically the PV of the expected cashflows that are expected to be paid over the life of the arrangements. We would contest that this is an arm’s length market rate. As [BNP] London would be taxable on any dividends received it will use a pre-tax discount factor when determining the appropriate value (ie 99.5 m). On the other hand, UK Investor will not be taxable on receipt of the dividends and would therefore use a post-tax discount factor when determining the price that it is willing to pay (ie 100m). In short, both prices should be regarded as arm’s length market prices. It is just the respective tax attributes of [BNP] London and UK Investor that give rise to different prices. With regard to the price that a UK group company will be prepared to pay for the rump of the shares in 2006 (ie 41m), this will be calculated to ensure that [BNP Lux] makes a return of £350k from its involvement in the arrangements. This is a significant profit and one that is justified by the level of risk that [BNP Lux] takes in the transaction (ie there is not a firm commitment by the UK company to purchase the shares at the outset of the transaction). With regard to the UK company, it will be prepared to pay 41m because it has been offered equity funding to acquire the shares and will therefore have no associated funding costs.”

[113]Mr Peters said in his witness statement that this email was an example of him providing Ms Majchrzak Gilot with indicative pricing of the various legs of the transaction for BNP Lux’s approval in the expectation that Ms Majchrzak Gilot would discuss it with the BNP Lux management team. He said that it was clear, therefore, that BNP Lux was aware that it was proposed that BNP Lux would receive a lower price for the sale of the dividend strip than the London branch would on the onward sale. He said he did not think this would surprise his colleagues in Luxembourg as they would appreciate the work that the London branch had done in conceiving, structuring, marketing and negotiating the transaction for which the London branch would expect to be compensated. He noted that the London branch would expect to pay less for the dividend strip than ALIL because it would have had to pay tax on the fixed rate dividends.

[114]On 2 December 2005, Ms Majchrzak Gilot sent an email to Mr Peters (and others) it appears querying the price to be paid by the London branch. She said: “to be clarify; sale price dv have to be = fair value of dv in T1 in Lux”. Mr Peters said he understood this to mean, in essence, that the price paid by the London branch had to be determined on an arm’s length basis. His view (as set out in the email of 24 November) was that the difference in the tax position on the fixed rate dividends for the London branch and ALIL justified the difference in pricing. He also noted in his witness statement that the dividend rate was unattractively low to the London branch. Although the London branch could deliver an acceleration notice requiring the transaction to be terminated early, there would be an exposure to this low rate during the three-day notice period. He also said in the statement that his understanding was that BNP Lux were keen to realise a pre-tax return from their involvement in the transaction and it was clear to him that this could not be achieved otherwise than through a sale of both the dividend rights and the shares in HIL (and his oral evidence on this point accords with this).

[115]In correspondence from 28 November to 2 December 2005, Mr Heres also raised a number of queries with Mr Williams as regards the arrangements for the sale of the “rump”. He asked whether it was really necessary for BNP UK to be funded “with a dotation from France rather than by debt” in order to acquire the shares from BNP Lux. The response was that it was necessary as:

“the strip shares pay no return for three years, as the rights to dividends have been sold [BNP UK] will not be able/willing to fund any purchase from debt. It is therefore very important that [BNP UK] is further capitalised to allow it to be able to make the purchase.”

[116]Mr Peters commented in his oral evidence that just like BNP Lux, BNP UK needed to be “commercially incentivised to enter into this transaction and in UK’s case it was a flat position … I mean one might suggest that after three years it would have the use of 60 million of cash that remained in [HIL] but by equity funding as opposed to debt funding, we were sure that [BNP UK] was economically flat.”

[117]Mr Heres also asked if it was alright for BNP UK to purchase the ordinary shares at a price below market value as he noted that the net present value was well below its purchase price. The response was that it was not alright as:

“[BNP Lux] has to make a profit from its involvement in the arrangements. Therefore should [BNP UK] decide to purchase the shares, they will have to do so at a price in excess of the net present value so as to give [BNP Lux] the desired and required profit. This is why [BNP UK] has to be funded by new equity from [BNP].”

[118]Mr Heres queried the response saying he did not follow the analysis. He thought that the price was above market as it was not calculated by reference to the fair value of the shares but in order to give BNP Lux its agreed profit. Mr Williams replied that “the way [BNP] funds its acquisition makes no real difference” to him; “it was a corporate governance issue”. He reminded Mr Heres that the reason that the £40 million was there was to ensure that the new company which issued the shares had sufficient monies to pay taxes due on its interest receipts. He noted that it would not be grouped with BNP for UK tax purposes and, therefore, had to pay the tax notwithstanding that the London branch would receive a large tax deduction. Given that this meant there would be no return from the shares for three years he thought this made it questionable from a corporate governance position unless the shareholder had free funds to fund the equity.

[119]Mr Heres continued in his email of 28 November 2005 to say:

“Anyhow, if you fund the acquisition by equity from France, the same question arises for [BNP]. What is the interest of [BNP] to make this investment?”

Mr Williams said that it was not the same question for BNP because BNP was, through its London branch, obtaining cheap funding with a sizeable pre-tax and substantial post-tax benefit from the transaction such that it had a clear commercial motive for capitalising its subsidiary to buy the shares.

[120]Mr Heres concluded by asking “I wonder whether it is not simpler and safer to have [BNP Lux] remain the owner of the ords.” Mr Williams thought that was possible on the basis that BNP could capitalise BNP Lux but that would be subject to further tax and accounting advice and he had thought it simpler to go with the other solution of BNP UK buying the shares through an equity funding.

[121]On 1 December 2005, Mr Scholes had a meeting with Patrice Pouliguen in Paris. The minutes of the meeting, circulated some time later, recorded the following as regards the description of the transaction again indicating that the purpose of the transaction was to obtain the s 730 benefit:

“This scheme involves the setup of a NewCo, which will pay a fixed rated dividend. [London branch] will acquire the rights to dividends and re-sell it to an investor (Alliance & Leicester). [London branch] will enjoy a full deduction of the purchase price, whereas the proceeds from the sale will not be taxed in the UK, provided the transaction can be considered as a trading transaction, which should not prove very difficult to demonstrate, as it is part of the [London branch’s] business to acquire assets to re-sell them shortly afterwards (even though these are assets are not usually stripped dividends).”

[122]In a further email exchange of 2 December 2005, Mr Heres again questioned how the rump was to be dealt with:

“As I already told Nick [Williams], the TCC approval will be sought on the grounds of the present structure, i.e. the sale of NewCo Ords by [BNP Lux] to [BNP UK]. Yet before this sale takes place, we want alternatives to be studied. We don’t believe the shares should be kept in Luxembourg as we don’t want to use Lux tax capacity. An alternative to having [BNP UK] buy the shares and be funded from France, would be to have [BNP] buy them directly. Would this seem to be feasible to you? On a UK tax perspective, would it be acceptable to have the shares within the same legal entity than the London branch?”

[123]Mr Williams responded:

“We will give this some further thought. We will plan to have a firm strategy within a couple of weeks, which may result in [BNP UK] considering the purchase of the shares in mid-January. Unless Peter has a different view, my understanding is the UK tax analysis is substantially unaffected if the UK is eventually UK owned or held elsewhere…… I understand that we would be slightly more comfortable if the shares were not held by UK company, hence our suggestion that the transaction be approved by TCC with the prospect of having the shares bought by [BNP UK], but with a commitment to examine between now and mid-January whether an even more satisfactory structure could be achieved. We can always strive for the truth but closing acceptable transactions takes priority … [He said Luxembourg tax capacity was not used and he thought it wrong to add in French tax capacity.] I cannot get too excited on this point as on any measure, this is a terrific transaction for [BNP].”

[124]It was put to Mr Peters that it was essentially certain at this time that BNP Lux would sell the shares to BNP UK. He said that he thought Gestion Financiere (“GF”) had asked if there are alternatives and there would have been alternatives, for example, BNP Lux could have received a fee or it could have been capitalised: “There are a number of ways we could have ensured that Luxembourg was commercially and economically incentivised to enter into the arrangements. So the structure or the idea that we thought was best was Luxembourg selling the rump possibly to [BNP UK], but that was not set in concrete. It follows all the papers because we wanted to do that. As we said, it was highly likely that we would deliver that. But, as I said, I seem to remember from the [GF] approval paper that they asked us also to look at alternatives.” (This was the case as set out below.)

[125]Mr Scholes was copied in to Mr Heres’ email and he responded as follows:

“My understanding was that if I had insisted on the shares being retained in Luxembourg, the transaction could not practically have been completed before the year end. This was the main reason for not doing so…… As I said before, I am prepared to sign off on the basis of a future sale of the stub, but given we are reconsidering this in the New Year, I will be very keen indeed for us to find a way for the original Lux owner to retain the shares as in my view this materially improves the UK tax analysis. I am concerned that this prospect appears to be receding in everyone’s minds. [If the shares are in fact sold, I do not think that [BNP] is an option.]”

[126]It was put to Mr Peters that the reason why Mr Scholes was vetoing this suggestion was that it would make it very obvious that BNP was subsidising BNP Lux to make the sale of the strip at an advantageous price to the London branch. Mr Peters could not remember the precise issue but thought that Mr Scholes’ concern was to do with the possibility that BNP Lux could be taxable on the strip proceeds if it sold in the same fiscal year.

[127]Mr Williams completed the email exchange on this topic by stating the following to clarify his understanding:

“We (OF) will do all we can to investigate other (non-UK) purchasers of the shares in Newco from [BNP Lux] over the next month or so. For the reasons we have discussed previously (ie Lux regulatory, Lux accounting, Lux tax existing holding group structure of [BNP Lux]), it has always looked and remains the case that it is very unlikely that [BNP Lux] will hold the shares indefinitely. Whilst nothing is predetermined, it is anticipated that sometime during 2006 [BNP Lux] will look to dispose of the investment in Newco shares. We (OF) will investigate between now and then in cooperation with our colleagues in [BNP Lux], prospective purchasers of such a shareholding. As promised we will work hard (with [BNP Lux]) to evaluate and research and all possibilities to locate a potential non-UK purchaser of the Newco shares. I assure you this is at the forefront of my mind. Pls let me know if I have misunderstood the situation, as I do not want any surprises or mixed messages at the TCC on Tuesday.”

[128]Mr Scholes replied that so far as he was aware there was no scope for surprises or mixed messages. He had always made the position clear as set out in his latest email.

[129]The OF team appear to have taken on board Mr Scholes’ concerns. On 2 December 2005, a member of the accounting team (Mr Trifiletti) sent an email to BNP Lux copying in Mr Bayfield stating that “your preference is not to report a loss on the investment and a subsequent gain in 2006. To avoid this situation, it’s proposed that the UK acquirer provides you in January with a statement of intent to purchase the entity at 41.7 million or at least 40.5 million”. Mr Bayfield responded:

“To come back immediately on your e-mail, one point which is out of the question is that BNP UK issues the letter giving a statement of intent to purchase the shares in [HIL]. It is crucial to our tax analysis there are no arrangements for the sale of ordinary shares in place, even a statement of intent.”

[130]On 2 December, Ms Majchrzak Gilot sent an email to Mr Peters querying whether the price paid by the London branch could be said to be an arm’s length price given that it was different to the price paid by ALIL.

[131]As noted, it is clear that the size of the transaction was determined by reference to the tax capacity of the London branch. There is an indication of this in an email of 2 December from Mr Robinson to Mr Heres (copying Mr Scholes and Mr Delafontaine). It seems to have been in response to a proposal by Mr Heres to increase the transaction size and asking the tax team for their view:

“in terms of your suggestion that we consider increasing the size of the transaction I can confirm that the client ([A&L]) are keen to do so. However, there are arguments both for and against increasing the transaction size. I have a slight preference to do so but would not resist a decision to stay with the current size. I have asked group tax in London (Peter Scholes) to consider this question and they will call you directly to discuss.”

Approvals process

[132]The transaction needed to be approved by GF and by the group’s Tax Coordination Committee (“TCC”) to go ahead. GF is a division of the group’s finance department (the others being accounting, management accounting, group development and financial information/investor relationships). Its activities included:(1) Managing mergers and acquisitions carried out by the bank.

(2) Managing the structure of the group from a legal, financial and regulatory perspective.

(3) Managing investments made by the bank in third companies to the extent that these were not allocated to a particular business line.

(4) Overseeing financial issues of an operational nature within the BNP group.

(5) Managing the tax position of the group.

[133]It was in relation to its tax function that GF was involved in the review of all transactions with third parties that have a significant impact on the tax position of the BNP group. GF was only ever involved in reviewing transactions referred to it by the various business teams within the BNP group; it was not involved in third party transaction origination.

[134]Once GF gives its approval to a transaction, it is then recommended to the TCC which gives the highest level of approval. At the time of the transaction, Mr Demon was a member of the TCC representing GF. The BNP group CFO, to whom Mr Demon reported directly, was also a member of the TCC. Mr Demon described the TCC as a coordination committee governing the tax and finance groups and topics of common interest to them. The TCC met fortnightly and he estimated that, at the period in question, it would consider around 12 transactions each year although it also had other business to deal with. He agreed that the reason the current transaction was referred to the committee was because there was a “very important tax element”. He explained that the Transaction Approvals Committee (“TAC”), which also approved this transaction, had a different function as a credit or business committee, which reviewed a much larger number of transactions in the order of 200 to 250 in each year at the time in question.

[135]Mr Demon explained that he had joined BNP in 1992, he had been in the GF finance team since 1997 and headed that team from 2003 to 2007 when he joined the Global Equities and Commodity Derivatives section of BNP in Paris which then merged with fixed income to create the current global markets business. His role at the time was essentially focussed on risk management on the credit side.

[136]The different functions involved in the transaction prepared papers, which were finalised in late November and early December, for the TCC’s consideration. These included papers prepared by each of(a) GF dated 2 December 2005

(b) the London OF team, Ms Majchrzak Gilot and Mr Berg dated 29 November 2005

(d) the group tax department dated 6 December 2005 and

(e) the group accounting department dated 29 November 2005. Mr Demon confirmed that he was involved in drafting the GF memo and that he also reviewed the other papers in his role as a member of the TCC.

[137]An outline of the content of the papers produced by GF, OF and the tax teams is set out below. In the approvals papers, other than the GF paper, whilst the tax benefit expected to be achieved by the transaction is acknowledged, the transaction is presented in headline terms as a financing opportunity. The email correspondence regarding the approvals also reflects this presentation. On 28 November 2005, Mr Williams circulated the document by email to some of those involved in the approvals process stating that:

“[OF] are currently in the process of proposing Project Sumatra for approval, which is £100 million 3 year fund raising from [A&L] at an attractive pre and post-tax costs of funds. I have attached our approval document to this email for your review.”

GF paper

[138]The GF paper focused on the tax and financial aspects of the arrangements.(1) Early in the note, the benefit of the expected tax advantage was acknowledged. It is noted that the benefit was obtained immediately: “the transaction is expected to last for 3 years, yet the [London branch] would benefit of the transaction immediately after implementation, which should be before year end”.

(2) At the end of the section describing the transaction there is a reference to the proposed sale of the shares by BNP Lux: “Early 2006, it is planned that [BNP Lux] will sell the ordinary shares in Newco to [BNP UK] at a price slightly in excess of £41 million. This UK holding company would be funded by a capital increase subscribed by [BNP] in France.”

(3) The effect of the arrangement was stated to be that, due to an “unintended effect” of the UK tax law, the purchase of the dividend strip by London branch would be tax deductible whereas its sale to a UK investor would constitute “a tax exempt revenue”. It was also noted that BNP Lux would not be taxable on the receipt of the price for the dividend strip and the UK investor (ALIL) would not be taxable on the dividends.

(4) In the economics section, it was noted that, for an investment of £100 million, the UK investor (ALIL) would obtain a return of 4.23% which is 101 basis points over the market rate (giving a return over 3 years of £3 million). It was stated that for Newco (HIL) the non-deductible dividends would give a pre-tax positive position of a rate of funding at 35 basis points less than the pre-tax market rate: “yet absent the tax mechanism, it would be post tax negative, as the dividend is not a deductible expense.”

(5) The committee considered the hedging and recommended that further steps were put in place to ensure that BNP Dublin was fully hedged as at that point there was no hedging of the “rump”.

(6) In the tax section, it was noted that, given the transaction would have to be disclosed to the UK tax authorities within five days of its inception, “it is likely that the law will quickly be changed, making repeat deals less likely”.

(7) In a section on the use of tax capacity it was noted that: “The operation will use the tax capacity of the [London branch] for an amount equal to the purchase price of the Dividend Strip. As Newco will not be part of the UK tax group, its taxable income should not be taken into account. In addition, the deduction of the funding will also use tax capacity mainly for [BNP], because of its funding of BNP UK’s capital increase. The yearly amount will be slightly less than £2 million, which is negligible. The most recent estimate of London tax department is a tax capacity for the UK tax group of in excess of £200 million for the year 2005. Therefore it would be in the group’s interests to have the purchase and sale of the dividend strip take place before the year end of the group.”

(8) In the section on profitability, it was concluded that over the life of the transaction, the BNP group would have cumulative net income after tax of £26.2 million, cumulative net income before tax of 0.3 million and an average return on tax capacity of 25% (taking into account capacity in the UK and France). It was also noted that the UK investor (ALIL) would have cumulative net income after tax of £3 million and finally that: “The tax advantage will be shared the following way 90% retained by [BNP], 10% paid to UK Investor.”

(9) In the conclusion, it was stated that the transaction has a good return on tax capacity again referring to this as being 25% and that the tax capacity of the UK group for year 2005 was anticipated to be well above that requirement. It was noted again that the transaction relies on the “unintended effect” of the drafting of UK tax law, and it should “be expected that the UK tax authorities may want to challenge it on other technical and anti-avoidance bases but although there are a number of lines of attack BNP have the better arguments on all bases”.

(10) OF recommended sign-off subject to no change in the opinions of PwC and Clifford Chance, that repeat deals were to be presented to the TCC, and that, before the implementation of any sale of the ordinary shares, “alternatives to the structure presented by the metier should be considered” and that any entity involved in the transaction should be properly hedged.

[139]It was noted to Mr Demon that the memo recognised that the sale of the shares by BNP Lux was “planned” and it was suggested that this was because the memo was candid in recognising the reality. Mr Demon said that this was merely recognising that was the plan of OF but that the conclusion of the GF (as set out in 138(10)) was that OF should consider alternatives.

[140]It was suggested to Mr Demon that this paper was very candid as it focused on the tax benefit (see 138(1) above). He said that the reason tax featured so largely in the memo was because that was the function of the TCC and the paper was to be presented to the TCC: “So the fact that we’ve focussed in this memo on the tax capacity or on the tax aspects in general again should mean no surprise, because it is part of a process leading to a TCC”.

[141]Mr Demon said he absolutely agreed with the description of the economics:

“Again, there are two elements which are clearly stated. This transaction, there’s clearly two tax features: one which is advantageous to [BNP], which is a deduction under s 730, and one which is disadvantageous to [BNP], which is the non-deductibility of the flows paid to A&L. As you said, the memo is very candid and just describes the facts, or those two facts.”

[142]It was noted that the memo refers to repeat deals not being likely due to the change in law risk (see 138(6)) and suggested that was inconsistent with Mr Demon’s view as to why there was no repeat transaction. He said that he thought at this stage there were no other transactions in the pipeline and “again, for them to be stopped, they should have started. I’m not sure they - they had started at that point.”

[143]It was put to Mr Demon that the reason the tax profitability was put at the top of the section because that was the real reason this transaction was being done. He said again that the focus on tax was because the TCC, for whom the paper was intended, was “interested about the tax aspects of the transactions. There are other committees interested in other aspects of the transaction, including the TAC”. In addition, he disagreed with the point that the fact that tax considerations were very present in this memo was a sign that this was all about tax and that there was no other element in the transaction.

[144]It was noted to Mr Demon that he had said in his witness statement that there is an unwritten rule that the group would not undertake transactions with no commercial purpose and, therefore, that it must be part of the function of the TCC to consider the commerciality of the arrangements. He replied that “yes I agree with that, the transaction the commercial aspect or the funding element of the transaction was quite obvious.”

[145]It was put to him that the whole framing of this memo, in terms of its analysis of profitability (by reference to tax capacity) and its conclusion (it should be implemented before the year end), was all focussed on the tax, and that was, not simply because it was a memo to the TCC but because the memo was candid as regards the real purpose behind the transaction. Mr Demon said: “I disagree with this. The TCC was not just - I think the memo was really addressing the points that the TCC would consider in making a decision about that transaction and the funding element, 150 or in the memo 100 funding is obvious. Maybe we could have written one page about it but I don’t see exactly what we can add. It’s obvious that this transaction provides funding of 100 million in the memo, 150 million as it was implemented … There is not much to say about that”.

[146]Mr Demon agreed that, looking at the above figures, the tax advantage that the group expected to obtain from the transaction “absolutely is massive” compared with the very small pre-tax element. OF approval paper

[147]The main points of note in the OF paper are as follows: (1) As noted, the transaction was presented as a fund-raising opportunity at an attractive rate.

(a) It was described in the heading as:

“An opportunity for the [BNP group] to raise funds £100 million at an attractive rate by selling a dividend strip to the Alliance & Leicester group”. (b) Similarly, it was described in the executive summary as having the aim of raising finance: “the aim of this transaction is to raise £100 million of floating rate general funding for the BNP group for 3 years at an attractive pre and post-tax cost of funds”. (c) The funding was described as being “tax efficient” for BNP due to the s 730 benefit and for ALIL due to the receipt of tax free dividends. (2) It was explained that, if executed before the end of the year, the transaction would result in an increase in the BNP group’s post-tax net income of around £29 million. (3) It was noted that nineteen investors had been contacted and four remained interested with A&L able to close before the year end. OF asked for generic approval for further transactions with the other three investors up to an amount of £150 million for 2006. (4) The tax benefit expected to be achieved by the transaction was explained further as achieving, at the post-tax level, a tax deduction of £98.9 million at the point of sale of the dividend strip but noting that the group will pay non-tax deductible dividends over three years of £4.37 million. Therefore, “this transaction will have an aggregate post-tax benefit of £25.8 million and will impact upon the consolidated post-tax earning” as regards the year 2005 of “£29.7 million” and as regards 2006, 2007 and 2008 of a “decrease of £1.3 million”. (5) It was noted that as “the [BNP group’s] benefit should be secured at the point that [the London branch] disposes of the dividend rights, the break-even point should be reached immediately”. (6) It was noted that the downside is that if the tax benefit is not obtained the group would have paid implementation costs (which at that point were in the region of £370,000 plus VAT) and paid dividends at an “enhanced post-tax rate” which was stated to give a maximum post-tax cost of £3.5 million over the three years. (7) The appendix to the memo set out the features of the transaction including that early termination rights were included but noted that: “although [OF] have negotiated free exit rights with [A&L], this very favourable position was negotiated in good faith on the basis that [BNP] will not exercise its rights unless there has been a material change of assumption (for example, a change in tax or accounting treatment). To exit for any other reason may seriously affect [BNP’s] group relationship with [A&L]” (8) The potential sale of the “rump” was also set out in the appendix: “It is possible that at some point in the future, perhaps early in 2006, [BNP Lux] will sell the ords to [BNP UK]. Should this occur, [BNP UK] would issue new equity to [BNP] for cash consideration in order to fund the 41.6 million purchase price.”

[148]Mr Peters was questioned on whether this was a fair representation of the plan as regards the sale of the “rump” given that he had confirmed that it was always in prospect and he had described it as “highly likely”. He agreed that the way in which this was presented in the approval document was slightly disingenuous:

“The genesis and evolution of this transaction was very quick. We needed to get to approvals quickly, so these documents were written quickly. There was a lot of thought and time and effort went into them, but they are a summary of what actually happens. That’s what we present to the approvals committees. This is probably slightly disingenuous language. As I said yesterday, we were going to try very hard to ensure that the rump was sold by [BNP Lux] and, by selling that, that they would achieve a profit from its involvement in the transaction. … By the time we submitted it, we were probably looking at [BNP UK] quite hard.”

[149]In re-examination Mr Peters said that to say the sale of the ordinary shares was “likely” was a more realistic appraisal.

[150]It was noted to Mr Demon that OF were asking for generic approval for further transactions so that it was clear that further transactions were contemplated. He agreed but said that he thought there was nothing really in the pipeline and so nothing to be stopped by a change in law.

[151]It was put to Mr Demon that the explanation of the purpose of the transaction in this document was disingenuous in stating that it was to obtain financing at an advantageous rate, as the real aim of the transaction was to obtain the tax benefit. He replied that the aim was both.

[152]Mr Peters confirmed that, as set out in the approval document, the group could have terminated immediately, but he thought that from A&L’s perspective this was a good investment opportunity and “they would have been very annoyed, had they gone to all the effort that they did in November and December to put the transaction on foot, had we terminated it after a few days.”

[153]It was put to Mr Peters that the costs incurred in relation to the transaction, which at the point the approval document was produced were £370,000 (and in fact totalled £500,000), were well in excess of the costs which would be incurred on a conventional loan. He replied that he suspected that was “absolutely right”. It was noted that the benefit would not exist at all unless the transaction went on for at least a year because the pre-tax benefit for the first year would be wiped out in full by the costs. It was put to him that there was a tension between that and the fact that the longer the transaction went on, the greater the eroding of the bank’s benefit. He agreed that if the trade had unwound early (given that it could be terminated on short notice) the costs would in effect have wiped out the expected pre-tax benefit for the first year but he said that was just a risk the business took in the same way that it took the risk of the trade not closing at all.

[154]Mr Peters also acknowledged that the group was limited in what use it could make of the funds raised from ALIL given the termination rights. The result was the BNP group raised “£150 million of funding for as long as that funding remained in place at a discount to Libor. So it was a daily accrual, if you like, of 38.5 basis points or 38.6 basis points benefit”.

[155]It was put to him that, given the size of the post-tax benefit compared with the pre-tax benefit, it was the post-tax benefit that BNP were trying to secure from this transaction. He agreed it is “the largest benefit. There were other benefits, the fundraising and the discount”.

[156]It was put to him that, looking at the effect of the tax on the income arising to HIL for each of 2006, 2007 and 2008, the net cost would be much greater than the pre-tax saving resulting from the discount over Libor. Mr Peters said:

“Yes, but the point of this lending transaction is that unless the bank is able to achieve a deduction at least equal to the cost of the non - the tax cost of non-deductible dividends, then it’s uneconomic from a post-tax perspective. But we were advised at the time that the 730 deduction would give us such a deduction and that was more than compensating the fact that the dividends were non-deductible. You know, in many [ways] it’s just akin to a vanilla loan - I mean, the numbers are larger, but if you don’t get a deduction for your interest on a vanilla loan, it becomes uneconomic. This, the deduction was much larger. It comes, perhaps, with some more risk of challenge, but, nonetheless, when we entered into this transaction, we thought the technical analysis was such that you would get a deduction and that deduction was in excess of the cost of the non-deductible dividends…”

[157]Mr Peters agreed that the risk of challenge is “obviously massively” different to that as regards a conventional loan.

[158]It was put to him that the cost of the dividend which was a maximum post-tax cost of £3.5 million was well in excess of the cost of the normal commercial loan from A&L. He said:

“I’m not sure that’s right…It is certainly the case that if there was no deduction secured equal to those dividends on a post-tax basis, it would have been uneconomic to make that loan …. Not necessarily well in excess, I would need to look at the numbers, but it would be uneconomic.”

[159]On the front cover of the draft approval document circulated shortly before it was sent to the TCC, there was a description of the project name used for the transaction of “Sumatra” as a wind which; “occurs in the north Indian Ocean at any time of the year, but is most common between April and November. The wind blows at between 20 and 30 knots, usually occurs at night and lasts for between one and four hours.”

[160]When he saw this, Mr Scholes wrote to Mr Peters (it appears somewhat crossly) on 2 December 2005 as follows:

“James, I can’t believe you issued this in final with a description of the wind. This will go to the Revenue eventually and it will look as though you are playing some sort of game. Will you please reissue the document without this.”

[161]Mr Peters explained this as Mr Scholes wanting to make sure that “we present these transactions in the best possible light”. UK tax approval document

[162]The memorandum prepared by the UK tax team included the following:(1) In the section headed “tax risk”, it was noted that the tax benefit under s 730 “appears to be an unintended effect of the legislation and we would therefore expect the transaction to be challenged on other technical and anti-avoidance bases.”

(2) It was noted that HMRC could be expected to act quickly to remedy this defect once it came to their attention (noting that it was intended to disclose the transaction to HMRC at an early stage) but “as London branch obtains the full tax benefit from this transaction within the first few days of execution”, barring a retrospective change, the tax position was expected to be secure.

(3) As regards change of law risk, it was stated that the transaction would be reviewed after the pre-budget statement on 5 December 2005 and the tax team would circulate a note by 9 December 2005 setting out whether the expected tax treatment of the transaction was affected.

(4) It was stated that the transaction should be disclosed as regards the s 730 benefit.

(5) It was concluded that overall the structure should work and was “in line with the commercial objectives of the BNP group and should result in both a pre-tax and post-tax positive return”.

(6) Under the reputational risk heading, it was noted that it was expected that HMRC would view this as “aggressive tax planning”. However, the UK team considered that, provided that the London branch filed its corporation tax return to reflect the correct technical position and complied with its disclosure obligations, there should not be any reputational issues in the UK. It was then stated that: “In particular the transaction is designed to achieve a strategic objective of the group (to raise funds at a preferential rate) and to provide the London branch with an opportunity to on sell a financial asset to an external client at a profit”.

(7) The recommendations included that, in the event of a sale of the dividend rights by the investor, OF should immediately take steps to accelerate the payment of the termination dividend and then wind up the transaction.

[163]In looking at this document, Mr Peters was questioned on the perception of the likelihood of challenge. He confirmed that his understanding was that there was “a high likelihood of challenge” and said that “it was clear that the Revenue would act very quickly to close this as it would have an unintended effect.”

[164]Mr Demon was asked whether he agreed that it was aggressive tax planning. He said it was aggressive in the way that a major tax benefit was achieved and it was an unintended effect of the law. It was put to him that it was also aggressive because the tax loss which arises in the London branch does not reflect the economic reality of what has occurred in the sense that the London branch had not lost £100 million in this transaction. He said:

“There is a discrepancy between the tax result and … the accounting result.”

[165]It was also put to Mr Peters that the references to the strategic objective of the group were disingenuous. He responded that to say that this was a strategic objective of the group “probably is a bit over-egging it”. He accepted that it was not strategic:

“We would like to present the commercial rationale of the transaction. It is understood it’s a very large tax benefit, but of course we would present it in the best possible way.”

[166]The global tax group paper included, similarly to the above, a statement that “Project Sumatra enables BNP group to raise a £100 M of general funding for 3 years at an attractive rate by selling a dividend strip to A&L”. TCC meeting

[167]The TCC met on 6 December 2005. Mr Demon could not remember the details of who was there. The correspondence shows the following as being present: Philippe Bordenave (Group CFO), Patrice Pouliguen (Group head of tax), Alice Aubert (the secretary), Jean-Vincent Massoni (Group head of transfer pricing), Mr Robinson, Mr Williams and Mr Peters. From the minutes of the meeting produced after the hearing it is also clear that Mr Heres and Mr Scholes were there as was Mr Levy-Garboua.

[168]Mr Demon did not appear to have much actual recollection of what happened during the TCC. His account from his witness statement and oral evidence is set out below. Following the hearing, the minutes of this meeting were produced and what these record is set out after Mr Demon’s evidence.

[169]Mr Demon said in his witness statement that the TCC considered the merits of the transaction by reference to the benefits it brought to BNP and the group on a before and an after-tax basis. The approval process, therefore, included an assessment of both the financial and tax elements of the transaction. He said:

“The transaction was ultimately approved because it brought financing to the BNP group (in the amount of £150 million) at a commercially attractive rate.”

[170]He went on to note that whilst the after-tax position was obviously an important factor in the transaction receiving the approval of the TCC, it was not the only aspect considered by the committee during the approval process. There was an unwritten rule that the BNP group “would not take part in artificial transactions, the sole aim of which was to obtain a tax advantage and which had no commercial purpose”. He stated that he had rejected a number of proposed transactions that did not comply with this rule when he was a member of GF.

[171]In his statement, Mr Demon stated that the TCC concluded that:

“in addition to the post-tax benefit the proposed deal achieved real and valuable commercial benefits for the Bank and the [BNP] Group as a whole, namely, the raising of financing from an external party [A&L] …on commercially advantageous terms. So whilst the TCC was of course aware of the Transaction’s post-tax benefit, it concluded that the Transaction was also commercially attractive on a pre-tax basis to both the London branch and the [BNP] group as a whole.”

[172]He went on to note the economic effects of the transaction essentially agreeing with the analysis of Mr Peters. He stated that, from a commercial point of view and from the perspective of the whole group, the purchase price of £150 million was equivalent to a loan which would be repaid by the termination dividend with the fixed rate dividends being economically equivalent to interest. He said that from a pre-tax point of view the profit to the BNP group arose as a result of the difference between the dividend rate being paid by the BNP group for the use of the £150 million, which was economically equivalent to an interest rate, and the higher interest rate which the BNP group would have had to pay in order to obtain the same level of credit for the same term in the market. The TCC took into account the negative feature, that the dividends payable by HIL were not tax deductible whereas payments of true interest would have been, and, that the London branch could treat the price paid for the dividend rights as tax deductible whilst not paying tax on the sales proceeds:

“Because the deduction available to London branch from the purchase of the dividend rights was broadly equal to the amount “borrowed” from ALIL under the transaction, from a post-tax perspective the Transaction was significantly superior to a straightforward borrowing by way of loan, under which there is no tax deduction allowed for repayment of the principal. It was obvious to us that the second factor…considerably outweighed the fact that the dividends were not tax deductible in [HIL]. Moreover, whilst a straightforward borrowing by way of loan would have been simpler to structure and execute than the Transaction, generally speaking the overall economic return on a transaction is more important to the Bank than the relative complexity of it. On that basis the Transaction was judged to be attractive as it was cheaper both in pre and post-tax terms than a straightforward “vanilla” borrowing.”

[173]It was put to Mr Demon that for his unwritten rule to be applied, the TCC would have had to consider the commercial aspects of the transaction and he was asked to confirm if this was the case. He said initially that what he meant in his witness statement was that the TCC considered the financial and tax elements by which he meant “essentially the profitability of the transaction and the tax considerations that we have discussed” but that was “not exclusive of anything else”. He then acknowledged that his witness statement referred to the unwritten rule that the group would not undertake artificial transactions, the sole aim of which was to obtain a tax advantage with no other commercial purpose, and he said:

“Now, yes, the commercial purpose is mentioned [in his witness statement] paragraph 13, and it was duly considered by the [TCC], but as I said before, this commercial purpose was obvious. There was nothing to write about it. It’s just - looking at the chart, you can see that there is - this transaction is a funding transaction from [BNP’s] perspective. Put aside the tax considerations.”

[174]Mr Demon went on to say that the rule was applied consistently over the years. He said that he had spent 10 years at GF so he knew what he was doing. He acknowledged that there was no reference to the unwritten rule in any of the memos prepared for the TCC, including the GF memo, but said he did not think “there was a need to express that rule. The memos are not about setting general principles or policies. It is just about reviewing a transaction.”

[175]Mr Demon was asked to clarify what would suffice to pass this test and if having a very small commercial purpose satisfied this test. He replied that this was a matter of judgement but, in any event, the raising of £150 million of finance over three years was not a small matter:

“In this case anyway, I would not consider that it was a very small purpose. We raised £150 million of financing from a third party over a period of – three years were contemplated, in the end it was two - a little bit less than two years. For me, this is not a very small element. This is a significant element. [BNP] needs funding.......this was a very significant benefit of the transaction to raise funding for [BNP]. [BNP] is not self-funded. We do not have enough deposits to fund our assets. So funding was indeed a very significant element in this transaction.”

[176]He was asked if it was alright for a transaction to have the main aim of obtaining a tax advantage and whether that would that pass the unwritten rule. He said:

“I guess so. It depends on the other advantage you obtained. It’s not a matter of one advantage being higher than the other; it’s a matter of is there - is there a substantial non-tax aspect in these transactions? In this case, I had no doubt, and I still have no doubt, that there was a substantial non-tax element in this transaction. £100, £150 million of funding over two years or three years is contemplated. It’s very substantial.”

[177]Mr Demon confirmed that he had also been involved in the approval of another scheme undertaken by BNP (being the scheme in HMRC v Fidex

[2016] EWCA Civ 385

[2016] STC 1920

(‘Fidex’)).

[178]It was put to him that given the TCC approved both that case and the current one, in each case plainly being tax avoidance schemes, the unwritten rule was virtually meaningless in terms of how it was actually applied, save as to refuse approval for only the very crudest of schemes. Mr Demon disagreed.

[179]It was put to Mr Demon that it was meaningless to claim that a transaction has a commercial rationale by reference to the before-tax position when, absent the s 730 benefit, it is loss making on an after-tax basis and, therefore, it could not be said that, in reality, the transaction provided cheap funding for the BNP group. He agreed that if one were looking at the transaction absent the s 730 advantage “it made no sense” and “we would never have entered into that transaction on that basis” but he thought the transaction should be viewed as a whole.

[180]The response was again that: “the commercial purpose is to borrow from [A&L] … if you can one step back… it’s not just [BNP], it’s a transaction between us and a counterparty. For me, it’s obvious that the common purpose of the parties to that transaction is for one to provide funding and the other to receive funding. I don’t understand what - why we can dispute that”.

[181]It was put to him that it must surely be funding on commercial terms, which this was not, in the absence of the s 730 benefit. He was asked whether he would approve a transaction where there was a provision of funding, but it was 5% above market rates. He said he would not approve such a transaction “because it would just be worthless to [BNP]. It’s not a question of commerciality … I would not approve it because it will make our shareholders lose money”. He said there can be a commercial rationale to borrow at a high price, but financially it would not make sense.

[182]It was put to him that the alleged commercial rationale in this transaction was illusory and was only possible if the s 730 benefit actually came good and that the real purpose was to obtain that tax advantage. He disagreed and said again that the commercial rationale was to raise funding from A&L:

“The specificity of that transaction is that instead of a vendor loan where the interest is deductible for the borrower and taxable for the lender in this case the interest is non-deductible/non-taxable and on top of it there is a deduction roughly equivalent to the principal of the loan. So this is a very special way of borrowing from A&L.”

[183]It was pointed out to Mr Demon that an unusual feature of the funding in this case was the early termination rights. He thought it was not so unusual and that feature does not disqualify the transaction from being a loan or economically equivalent to a loan. He noted that the transaction was recorded in BNP’s financial statements as a borrowing.

[184]It was put to him that, in particular, the funding was not much use in terms of capital raising as the BNP group were at risk of having to repay it at very short notice. He said that was not that unusual in loan relationships or lending relationships but agreed that it was probably unusual in a bond relationship. He also noted that it was “obviously in the interests of [A&L] not to terminate that financing, so from an economic perspective, we can say that we could count on the financing remaining in place, as it happened to remain in place during two years…. It was contractually possible, but economically … the interests of [A&L] was to continue the loan rather than discontinue it or terminate it”. He agreed that it was unusual for the borrower to pick up the lender’s costs.

[185]It was put to him that the transaction could not really be compared to a straightforward borrowing because this wasn’t really a borrowing transaction but a tax avoidance scheme, and what BNP were weighing up was whether the tax avoidance scheme was defensible. Mr Demon did not agree:

“From our perspective and from the perspective of our counterparty, again, it was a funding transaction. I do not dispute the fact that there is -it was also designed to achieve a certain tax result. It was a borrowing transaction represented as such in the accounts of [BNP], which were perfectly correct. I frankly, I don’t see how we can dispute the fact that this was a borrowing transaction. If you say so, you are implying that the financial statements of [BNP] during those years were wrong.”

[186]It was also put to him that this was not properly regarded as a capital market transaction; rather, this was a structure devised and negotiated by either tax professionals or people with a tax professional background. It had nothing to do with capital markets. Mr Demon agreed that it was structured by people with a tax background to a large extent but that “does not preclude or prevent the transaction from being a capital markets transaction”. From the perspective of the London branch, this was the purchase and on-sale of the cash flow and, from this perspective, it is typical of a capital markets transaction.

[187]It was noted that whilst he had maintained or held on to the commercial rationale, he had not denied that the transaction had at least a tax avoidance purpose. Mr Demon said:

“It has the purpose of initiating a tax deduction, essentially.”

Later he clarified that he accepted that obtaining the s 730 benefit was “one of the main purposes of the transaction”.

[188]Mr Demon was very firm that there was an unwritten rule of the type he described which he was very much aware of as the head of GF. We accept his evidence that there was such a rule and that he personally may well have had that rule in mind in approving the transaction as head of GF and in participating in the TCC. We accept that he considered that the requirements were satisfied because, as he said was obvious, the group obtained £150 million of funding and funding was always useful to the BNP group. There is some support for the fact that there was an awareness within the BNP team of such a rule or policy from the approach of Mr Peters. He noted in his evidence the need for the transaction to have sufficient commerciality both as that was what was needed to obtain the tax benefit but also as otherwise “these things would not get done”.

[189]However, there is no real evidence that the obtaining of the financing featured in the minds of those involved as the sole motivation for the proposal as we discuss in our conclusions. That includes the other members of the TCC as was apparent from the minutes. Nor do we think it can be said that the purpose or aim of the transaction was anything other than to obtain the s 730 benefit as that aim or purpose is to be discerned as an objective matter from the facts and the characteristics of the transaction.

[190]Mr Demon suggested in his oral evidence, in effect, that it was obvious that the purpose of entering into the transaction with ALIL was to obtain that funding. We feel there is some confusion here, however, between purpose and effect. The funding was obtained in the sum of £150 million and, therefore, that was clearly a commercial effect of the transaction. However, to our minds, it is not at all clear that this means that obtaining funding of £150 million was the purpose of the transaction. Rather the correspondence, memos and minutes as well as the nature of the transaction itself suggest that the funding was obtained solely for the purpose of enabling the BNP group to generate a dividend strip transaction in a way which, as Mr Peters said repeatedly, was designed to ensure that each group entity was commercially incentivised to enter into the transaction thereby giving the structure the commerciality the team thought was required for London branch to have the best chance of obtaining the s 730 benefit. In our view, overall, the evidence indicates that obtaining the finance was not an end in itself but rather a means to securing the availability of the s 730 benefit. We have set out our further thoughts on this in the discussion section as regards the trading issue.

[191]Mr Demon confirmed that, although this was not mentioned in the GF memo, the TCC had decided to increase the size of the transaction by 50% as they thought it was a good transaction. He thought it was not included in the GF memo because it was not a technical matter but rather a topic for discussion; it was a conclusion which would have been reached after the technical analysis was concluded. That the TCC did make this decision was confirmed in the minutes of the TCC which the tribunal received after the hearing. Minutes of the TCC meeting

[192]The minutes of the meeting are relatively short and were provided to us in an agreed translation in English from the original in French.

[193]The meeting commenced with Mr Scholes explaining how the tax benefit would arise. He then noted the regulatory position in Luxembourg and that BNP Lux would have to sell the shares promptly but that the sale would not be to the same entity as purchased the strip and it would not be until January 2006. The TCC raised certain specific aspects: the Luxembourg regulatory position, why a Cayman Island issuer was required, what the effect of the representations were to be given to ALIL, whether A&L properly understood the nature of the transaction as regards the effects for BNP and the impact of a change of law. Otherwise there is no indication that the TCC had any concern other than the availability of the s 730 benefit and maximising that benefit (through the increase in the size of the transaction), maximising the prospect of getting the benefit (through the timing of the transaction) and whether it could be repeated.

[194]Overall, there is no support from the record of the meeting for the assertion that the purpose or aim of the TCC in approving this transaction was to obtain funding from ALIL at an attractive rate. We note that, of course, the TCC had the various approval documents which to some extent in headline terms asserted that this was the aim of the structure. We accept that it was obvious from these that the group was obtaining funding under the structure at an attractive rate, assuming that the s 730 benefit was obtained (thereby more than cancelling out the negative impact of the non-deductibility of the fixed rate dividends by some considerable way). We note that there was no discussion or acknowledgment recorded that there was any attraction for the BNP group in receiving cheap financing from A&L under the proposed transaction and no indication that this played any part in the decision to approve the transaction. There is no record of any discussion of the commercial merits of the transaction or of any weighing of the potential disadvantage of the non-deductibility of the dividends compared with the potential s 730 benefit. There is no record of any discussion of the unwritten rule referred to by Mr Demon. We are unable to draw any conclusion that the TCC had this in mind when approving the transaction. Correspondence after TCC

[195]That the size of the transaction was to be increased was reported in email correspondence. In an email of 6 December 2005 (at 11.00pm after the TCC committee meeting), Mr Bayfield said:

“I am pleased to advise that the project received approval at TCC earlier today. The Committee recommended that the size of the investment to be sold be increased to 150m, an amendment which has been accepted by [A&L].”

[196]The increase was confirmed in an email from Tom Woulfe, of the Dublin branch, to Francois Van Den Bosch:

“New figures for Sumatra. [GF] were so impressed they asked if the deal could be increased by 50 per cent.”

[197]The next step was for the transaction to be approved by the TAC and their meeting took place on 8 December 2005. Shortly before the TAC meeting, Neil Robinson (the London Head of OF) apparently emailed Denis Autier and Christophe Delafontaine (Global Head of OF and Chair of the TAC) to share his thoughts on the development of the “business” in the UK: “…2. In an uncertain environment where disclosure can lead to a change in law it is critical that we try and source principal deduct transactions (such as Sumatra) when looking to optimize the tax position of the Bank and our clients in the UK. This will be difficult to achieve but we cannot afford to rely upon accrual transactions only.

[3]We should continue to work closely with the Group Tax Department (London) in order to manage the tax position of the Bank in the UK as effectively as we can – generating a balanced portfolio of accrual and principal deduct transactions to accommodate this objective.”

[198]It was put to Mr Peters that it was clear from this that Mr Robinson saw this transaction as a means of getting a deduction for the outlay and not as a strategic fund raising transaction. Mr Peters said:

“Well, from the point of view of what I can remember, my experience, was this transaction was known to have a significant tax effect or potential benefit to it. But that was never enough in the bank to do these transactions; they have to have strong commercial drivers. And this is a fundraising transaction by [BNP] from the [A&L] group…… Yes, the tax advantage is very important, but the transaction has commercial merit outside the tax and that is important for getting these sorts of things done.”

[199]On the day of the meeting, there was also correspondence regarding the credit for the transaction to be allocated to OF. Mr Peters said that he could not remember what was agreed here but he explained how the remuneration process worked. He said that, as for all business lines in BNP, the starting point for remuneration for teams is an allocation of the pre-tax reported profits earned. In transactions where there is a tax effect or tax enhancement such as in this transaction, financial management would allocate an analytical fee to the business to recognise a portion of that benefit, and that fee was often determined by negotiation.

[200]The correspondence indicates that the benefit to be allocated to OF for this transaction was 90% of a 75% benefit. The 75% reflected the provisioning which the UK tax team thought appropriate according to the level of risk of successful challenge. We note that the 90% corresponds to the 90% of the tax benefit identified in the GF memo on the basis that, in effect, the other 10% was provided to A&L.

[201]Mr Peters also reported the approval to members of OF on 6 December 2006 and noted the “need to find investors for repeat trades in early 2006.” TAC meeting

[202]Later on 8 December 2005, the TAC meeting was held in London at 16:15pm. The members of the committee were drawn from the various groups involved in the transaction. It considered the same OF approval document as had been presented to the TCC but with an updated appendix showing the steps involved in the transaction.

[203]The final version of the minutes of this meeting was circulated by Mr Bayfield on 22 December 2005. This contains some differences compared with an earlier version as explained below. (1) Both versions recorded that:

“the transaction represented an opportunity for London branch to make a significant profit (around £500K after costs) with minimal risk by buying and subsequently selling a 3 year strip of dividends in [HIL]. It was also explained that the transaction had certain tax benefits for both BNP and the client (A&L).” (2) Both versions contained the following as regards the increase in the size of the transaction: “It should be noted that the size of the transaction has recently increased from 100 million to 150 million. It was asked whether this was at the request of the client, and [OF] confirmed that this was the case.” (3) However, the earlier version had a statement that A&L had requested a much larger increase but had accepted a smaller one: “the client had actually requested an increase to GBP 450m but had accepted a smaller increase and agreed to discuss possible further business in the New Year”. (4) The earlier version stated: “Tax asked whether the shares can be sold to other companies as well as internal [BNP] ones, it was confirmed that this is the case.”

The later final version had a fuller description as follows: “Tax asked whether the shares in [HIL] may be sold to other companies in the [BNP Group] (perhaps outside the UK) as well as to UK subsidiaries. It was confirmed this was the case. [OF] confirmed that no negotiations would be commenced with any prospective purchasers until early 2006.” (5) The earlier version recorded that “it was asked if the board of [BNP UK] should be approached regarding the transaction. [OF] said that they should not be approached”. In the later version, the question was posed as to whether BNP UK had been approached as the prospective purchaser of the shares and the answer was given that they should not be approached until January 2006. (6) Both versions had a note at the end which stated that the original approval was given on the basis that at some point in the future, possibly in early 2006, BNP Lux may sell the shares to BNP UK and that subsequently it was decided that BNP Lux may be given this right of sale through the grant of a put option exercisable from 6 April to 20 April 2006. It was noted that the various BNP functions had been consulted and had approved this change.

[204]It was put to Mr Peters that there was never any plan to sell the shares in HIL outside the group as was indicated in the earlier draft of the minutes of the TAC. He said:

“Yes, I think that’s fair. And like I said a number of times, it was always - from our approval documents, it was always intended, or it was identified early that a suitable purchaser might be [BNP UK], and that we would work very hard to deliver that for [BNP Lux]. That’s why you see this in our approval papers so much.”

[205]It is clear, therefore, that whilst the plan was not to approach the directors of BNP UK until 2 January 2006, there was never any realistic plan to approach any other purchasers of the shares.

[206]As regards the size of the transaction, he agreed that there was a maximum deduction that “London tax would have allowed us to derive from this transaction, and that, to a certain extent, has driven the transaction size”. The “most probable reason” that the TCC was interested in doing a transaction at £150 million but not at £450 million in the 2005 accounting period, was because there was not sufficient tax capacity for the larger amount.

[207]On that day, Mr Bayfield circulated an updated appendix to the approval document showing what he said was the final fact pattern and reflecting the increase in the size of the transaction, which he said A&L had requested, to £150 million and accordingly £210 million as regards the investment required to be made by BNP Lux. Pricing of the purchase of the dividend strip and sale of the “rump”

[208]Shortly after this meeting, the pricing was settled. On 7 December 2005, Oliver Read circulated a computational model apparently using a discounted cash flow basis. This shows the price to be received by the London branch as £150 million and the sale price to be paid to BNP Lux of £148,475,000. This shows a pre-tax profit in London branch of £1.13 million but a substantial loss for BNP Lux as at 31 December 2005 of £1.783 million. Mr Peters confirmed that he had not been directly involved in the pricing but he had sight of the various financial models prepared in the period leading up to closing of the transaction. He recalled that, previously and at this stage, the modelling was done using a discounted cash flow method.

[209]Mr Read prepared further models the following day which were discussed with Mr Peters. In relation to the first, he stated that he believed that the strip price (£148,475,809) was also arrived at using a discounted cash flow calculation. By contrast, the price for the dividend strip on the second (£149,104,000) was arrived at using “goal seek”. This is clear from Mr Read’s email in which he circulated the second model where he said: “Apply two Goal seeks and get the following now”. Mr Stanton explained in his report that “goal seek” is a financial modelling function, available in Excel spreadsheets, that calculates values such that a desired outcome or goal is achieved. It is not disputed that the “goal seek” method was used to achieve two desired results: a pre-tax profit in London branch for £500,000 and an overall profit in Luxembourg from the transaction of £750,000. As regards calculating a profit of £500,000, for example, under “goal seek”, the set items would be everything except the acquisition price payable for the dividend strip to BNP Lux. The financial model would then calculate what the acquisition price needed to be to result in a £500,000 profit.

[210]Mr Peters did not know why the methodology had changed but he suspected it was by negotiation between the London team, the Luxembourg team and probably BNP in Paris. It was put to him that the reason why the methodology changed was that the less BNP Lux received, the more expensive it was going to be to buy them out in the end through the sale of the rump because their funding costs would be greater. Mr Peters agreed that:

“A feature of this transaction was Luxembourg was to make a profit, so it follows that … the lower amount…… it received for the strip, the more it would have to receive on the rump.”

[211]Mr Peters said that, although he was not involved in the pricing discussions, he thought that there was negotiation between London and Luxembourg about the overall level of remuneration that Luxembourg wanted to earn, and that would be factored into pricing, and the sort of bid-offer spread that London would want to earn from this transaction. The London branch had fixed a sales price of £150 million. It knew that it needed to purchase this strip for less than £150 million. This was worked out to ensure the agreed amount of profit for BNP Lux.

[212]Mr Demon was also questioned on why the London branch only received £500,000 whereas BNP Lux got £750,000. He said initially that it was a matter of good negotiation by BNP Lux. Later it was put to him that the reason London branch only got £500,000 was because London branch was not particularly interested in the amount or profit allocated to it because it had its eyes on the much bigger prize of the tax deduction and, therefore, it was willing to give £750,000 to Luxembourg. He said he was not involved in the discussion, but, “yes, I would agree that the tax advantage initiated by the London branch in the transaction was taken into consideration in the negotiation that happened between [the London branch] and [BNP Lux]”.

[213]On 12 December 2005, Mr Bayfield emailed Ms Majchrzak Gilot with the final figures for the sale of the dividend rights and a forecast price at which BNP UK “may be willing” to enter into negotiations to purchase the ordinary shares subject to internal approvals. The model was again a “goal seeking” one and it reflected the actual figures used in the transaction. The model shows a profit for BNP Lux of £750,000. Mr Bayfield said that this reflected five inputs: the investment of £210 million, the advisors’ costs of £150,000 plus VAT, the funding costs (which it is noted were variable), the price for the sale of the dividends and the expected strike price for the sale of the shares to BNP UK. He gave an indicative figure for the sale of the ordinary shares and noted that it would be calculated to ensure that the expected profit was achieved. He noted that the key variable would be interest rates and suggested that it should be calculated in early 2006 by reference to a formula based on the total costs incurred by BNP Lux to date (being the amount invested, the advisory and funding costs) plus an amount to allow for interest from the date the option was granted until exercise.

[214]In the days before the transaction was implemented, there was correspondence on the put option arrangement. On 9 December, Mr Peters wrote to Mr Berg confirming that it was envisaged that the option would be granted in January 2006 with exercise in April 2006. Mr Peters referred to an earlier call and said:

“For the avoidance of doubt, it was envisaged that this put option would be granted to BNP Lux for the first week of January 2006 and we would have an exercise date of 6 April to 20 April. The strike price would be agreed on the grant of the option.”

[215]Mr Scholes, who was copied into the email, responded that that would work from a UK tax perspective. Mr Berg then replied to Mr Peters querying why the transaction timing had to change and indicating he had thought that the put option arrangement was to be implemented at an earlier stage. He said:

“Could you please also confirm by mail the reason why you have to change again the deal (fiscal year, if I had well understood). We have yet received the different internal approvals to implement the transaction under the former agreed structure. So with a change of structure, we need to receive at least a verbal prior approval by our regulator.”

[216]Mr Peters responded:

“Broadly, the UK tax analysis looks at the situation where a shareholder sells the right to receive dividends without selling the shares. There is a technical concern that if the shares are sold in the same chargeable period as the dividend rights then this would not be helpful from a UK perspective.”

He went on to explain that for this purpose the chargeable period ends on 5 April and that Mr Scholes would not approve the transaction on the basis of the two things happening in the same period and that the grant of the put option in 2005 was also not acceptable.

[217]Mr Peters recalled that the reason BNP Lux wanted an earlier sale of the rump was that further approvals would be required from the Luxembourg regulator if, broadly, BNP Lux were to hold the shares for more than three months. He agreed that there was a tension between BNP Lux who wanted the exit to occur in January 2006 and the UK tax team who advised against an exit until after 5 April 2006.

[218]It appears that the views of the tax team held sway as, on 12 December, Mr Bayfield wrote to Mr Delafontaine, Mr Robinson and various other members of the finance function confirming the put option proposal had been approved by the TAC:

“We would like to advise you the TAC committee has now anticipated in early 2006 BNP Luxembourg may be offered the opportunity to acquire a put option over the ordinary shares from BNP UK, giving BNP Lux the right to sell the shares on the following terms, at the strike price to be agreed on the grant of the option expected to be determined on the basis that will give BNP Lux a profit from its involvement in the transaction after taking account of all expenses.”

[219]On 12 December 2005, Mr Heres confirmed to Mr Bayfield that GF did not object to the proposal for the put option arrangements:

“On behalf of [GF], I can confirm we do not object to this new structure, provided it is certain that the regulatory status of Luxembourg can remain as originally contemplated. Sale of the ordinary shares by [BNP Lux] would address this need. [BNP UK] could, with a funding from its mother company, be the purchaser. Further to indications by GTD, direct purchase by a French company should be ruled out.”

[220]Also on 12 December 2005:(1) Mr Bayfield emailed the TAC notifying a revision to the transaction structure. This email enclosed the post-script that is seen on both versions of the TAC minutes.

(2) Mr Williams sent Ms Majchrzak Gilot an email enclosing a further updated version of the model. This was to prove to be the final version and showed BNP Lux’s advisory costs of £150,000. The email also showed what OF intended to be the terms of the put option with BNP UK (or another BNP group entity).

[221]On 13 December 2005, Mr Williams sent the London Tax Department two emails (one being a redraft) concerning the pricing of the transaction with A&L. This addressed the 80:20 split of the “benefit” with A&L and appears to have been raised in the context of DOTAS concerns from the perspective of the tax advantage being obtained by A&L. He stated:

“As you are aware, the strip of dividends is to be sold to the A&L group by BNP London based upon an approximate 80/20 split of the benefit of the transaction in A&L’s favour, where such benefit is agreed between the parties to be the receipt of a tax exempt dividend by A&L. From my experience in the (tax exempt) investment market, pricing in this range is on market for the investor (A&L). Anecdotal evidence of transactions with comparable risks that have been concluded in recent month were undertaken, I understand, in the 75/25 range. In my opinion, it would therefore be very difficult for anyone to argue that the BNP group is receiving “premium pricing” from this transaction. A&L have been offered this pricing to reflect the significant relationship between our two organisations, our desire to undertake further business in the future, the relatively small size of the investment and A&L’s speed of execution and early commitment to this transaction. As you are also aware, we have approached other potential investors to consider this opportunity, one of which has expressed a serious interest to transaction early in 2006 at similar pricing.”

Implementation of purchase and sale of dividend strip

[222]On 13 to 15 December 2005, the transaction for the purchase and sale of the dividend strip was implemented as set out above. Correspondence in period leading up to the sale of the rump

[223]In the period following the purchase and sale of the dividend strip, there was a focus on the operational and economic aspects of the transaction in terms of the booking and working towards putting the put option arrangements in place.

[224]On 15 December 2005, Mr Trifiletti of the accounting team emailed Mr Williams and Mr Bayfield asking for their input on the booking of the sale and purchase of the dividend rights and whether what he proposed was acceptable. Mr Williams forwarded it to Mr Bayfield asking him to evaluate and sort it out. Mr Williams then sent a second email stating:

“Just make sure London makes a profit whatever you do.”

[225]Mr Read circulated a number of emails with explanations and/or models illustrating the financial position. On 22 December 2005, he sent an email explaining that the fees for A&L’s advisers were being paid by BNP but that BNP were getting them back pro rata but without compounding over the scheduled three year term of the deal by virtue of an increased spread in one of the components of the transaction. He said that the “arrangement ensures that A&L does not make a loss amounting to their legal costs if we were to exit the transaction early.”

[226]On 19 December 2005, BNP submitted its DOTAS disclosure to HMRC. It was put to Mr Peters that the transaction would not have been disclosable if the obtaining of the tax advantage were not, at least, a main benefit of the transaction. Mr Peters agreed. He was asked if he was, therefore, accepting that the obtaining of the tax advantage was regarded as, at the very least, the main benefit of the transaction. He said: “It was one of the main benefits, yes,…. that we raised funding from [A&L], and there was a very large tax benefit that came off the back of it”. He agreed that this “dwarfed” the other advantages.

[227]Mr Peters was also taken to some correspondence on the potential application of s 703 ICTA. It was put to Mr Peters that the UK tax team had considered the potential application of s 703 ICTA but did not appear to have thought of relying on the motive defence that the transaction did not have as one of its main objects the obtaining of a tax deduction. He was asked if that was a fair characterisation of their view. He said: “Yes, I think it would be - with such a large deduction”. It was put to him that the main object was obtaining that deduction and he said: “Yes”.

[228]On 6 January 2006, a member of the accounting team sent an email to Mr Trifiletti and Mr Williams stating that the net assets of HIL as at 31 December 2005 were £60,009,622.90. On 10 January 2006, Mr Heres sent an email to Mr Delafontaine and Mr Robinson attaching a draft accounting memorandum which noted that BNP UK and BNP would have to “write off part of the 62,3MGBP investment in [HIL] as it will only be worth £60MGBP”.

[229]On 10 January 2006, Mr Heres produced a further memo on behalf of GF. This included a section headed “treatment of the tax advantage of the transaction in the management accounts”:

“After inception and until the payment of the termination dividend to UK investor, the group bears tax on the income of the total of 210 million of assets of [HIL] whereas it gets a deduction only on the carrying costs of the net equity invested in [HIL] ie 60 million. The difference entails a tax cost of 4.74% x150 x 30% ie 2.133 million per year.”

[230]Mr Demon agreed that the point being made here was that the tax deduction for interest on the transaction would only cover in effect £60 million of the assets in HIL. As it was receiving taxable income on assets of a total of £210 million, the income on the balance of £150 million was essentially uncovered by any interest deduction on a group basis. He said that this was “another way of saying that the £150 million of funding that we got from [A&L] was not tax deductible. That’s exactly the same.”

[231]The memo continued to note that “it is assumed in this section that [BNP Lux] will exercise its put option and that the entity granting the put option will be BNP UK”. It was put to Mr Demon that it was always going to be the case in terms of practical certainty that a put option was going to be granted and that it would be exercised. He noted that the GF memo had required the “metier”, meaning the business line within the BNP group, to consider other options and, on the date of the TCC, there was no certainty that the shares would be sold to BNP UK. It was put to Mr Demon that there was no real attempt to look at alternatives. He replied that he thought it was investigated because the TCC was a serious body so that if that condition was written then it should have been investigated but he was not involved very much after the approval process.

[232]It was put to Mr Demon that earlier in the memo it is stated that:

“Early 2006, it is possible that [BNP Lux] will enter into a put agreement, whereby it will be able to sell the ordinary shares...”

It was put to Mr Demon that this was a slightly disingenuous view given that the memo then proceeded on the assumption that this would happen. Mr Demon did not agree; he thought the memo merely set out the position assuming that the option would be exercised if it were granted but was not assuming that it would be granted.

[233]It was put to Mr Demon that his view was unrealistic because, in early January 2006, Mr Williams and the OF team were putting together the proposal as regards the put option arrangements for the BNP UK board. In an email of 4 January 2006, Mr Williams sent the board of directors a paper stating:

“In advance of the board meeting of [BNP UK] tomorrow, please find below an outline of an investment opportunity to be considered by the directors.”

The memo noted the details of the proposal including that:

“In order to ensure that BNP UK is entirely protected in respect of a written put option, it is proposed that [BNP] enters into a conditional subscription agreement ...”

[234]Mr Demon agreed that it was pretty clear that, at this time, the route which was being pursued was a put option with BNP UK. The memo also referred to the need for London branch to have tax capacity. Mr Demon said that his understanding was that the transaction was producing a tax deduction and, for the London branch to benefit from the tax deduction, it needed to have tax capacity so as to offset the deduction against that capacity. He confirmed that the transaction would not have gone ahead if the London branch did not have the required tax capacity: “Yes, you can say that the tax capacity was a condition of doing the transaction, otherwise a significant element of the transaction would have been worthless to the branch”.

[235]He agreed that he would not have approved the transaction if there was not the required tax capacity and if the group did not get the perceived s 730 benefit:

“I agree with that because if this had been the case, then the transaction would have generated a net loss for [BNP] after-tax.”

[236]It was pointed out to Mr Demon that it was recorded in a note of a London/Paris tax department meeting of 18 January 2006 (at which, among others, Mr Pouliguen and Mr Scholes were present) that:

“Negotiations are still ongoing on for Project Sumatra. The tax audit team have stated that they think the transaction is very aggressive and that it will go to court. If the transaction does go to court, they consider that we will lose.”

[237]It appears that this is a reference to the tax team of the group’s auditors, as Mr Demon confirmed that the group do not have an internal tax audit team independent of GF. Mr Demon was asked if he was made aware of this opinion as head of GF and as a member of the TCC. He said he could not remember if he was told but he thought such a matter should have been brought to his attention.

[238]He was also asked if, when he was involved in preparing the GF memo, and when he gave approval to the transaction as a member of the TCC, he knew that the kernel or the core of this scheme had originated from external third parties, Swiss Re and Dresdner Bank. He said he did not remember but he was “not sure it would have make a huge difference anyway”.

[239]Mr Demon was asked if he knew that this scheme was, in terms of the s 730 mechanism, essentially being pitched by other financial institutions at around the same time. He said he knew that it had been proposed “by our own team [OF] to a number of counterparties and not only [A&L]”, but, he had “no recollection of being told about what other financial institutions were doing”.

[240]Finally, he was asked if he was told that “the supposed commercial rationale for this transaction was only devised after BNP had learnt of the s 730 benefit and in terms of how the scheme got developed” or in other words that BNP learnt of the tax advantage first and then came up with the commercial rationale second. Mr Demon said that “I think it was presented to me as a whole, otherwise the transaction would not have gone very far without a commercial rationale, it was my understanding that [OF] would I guess present me the transaction with all its components, not just a mere tax idea about using section 730”. When pressed, he concluded:

“I guess in the end, the genesis of a transaction is one thing, the reality is another one. In the end, if OF had presented me, let’s say, in two steps, because that’s what you imply, the same transaction, I would probably have made the same decision. As long as there is a commercial rationale in the end, I would have approved it in the same way.”

[241]In this context, Mr Demon was shown Mr Robinson’s email of 26 October 2006 and a draft transaction approval document. The point was made that Mr Robinson focused on the advantage of the s 730 benefit and noted the need for London branch to make a profit only as a tax requirement, in other words as a condition for obtaining the s730 benefit; it was not driven by some other commercial requirement. The approval document in headline terms presented the proposal as “an opportunity for the [BNP] group to raise low cost funds from selling a dividend strip to UK investor.” It was put to him that this illustrated the point that it was the s 730 benefit which came first and the commercial rationale later; the email of Mr Robinson reflected the real objective. The language used in the approval document was presented, it was suggested, with an eye on HMRC or possibly GF looking at it.

[242]Mr Demon said he thought that the funding element was implied in Mr Robinson’s email from the reference to needing a UK investor. So whilst “the presentation is different, but the reality is the same … I am just saying that those are two different ways to present the same reality, that the reality between the e-mail and the presentation is exactly the same.”

[243]It was put to Mr Demon that the purpose of the transaction, the reason why London branch was buying these dividends, was to obtain a tax advantage and that was fairly clear from Mr Robinson’s email. Mr Demon said: “That’s fairly clear from the e-mail, absolutely, but that does not exclude the commercial purpose of raising funding. On the contrary, what I am - telling you, I am trying to telling you, if it’s not clear, is that that very purpose is embedded in the e-mail”.

[244]On 12 January 2006, an email was sent by the Head of Statutory and Regulatory Reporting to various recipients including Mr Williams. The email addressed provisioning in respect of the Sumatra tax advantage but also identified “Sumatra 2” as well.

[245]On 20 January 2006, HMRC gave a press release indicating that the law on s 730 would be changed with effect from 20 January 2006. This removed the possibility of future Sumatra transactions occurring.

[246]On 24 January 2006, there was an email chain that discussed a request by BNP UK for a letter of wishes to be provided by BNP as regards the option arrangements. There was a disagreement between Mr Heres and Mr Bayfield about whether this should be given and Mr Heres commented that it would have been desirable to have the arrangements with BNP UK agreed at the outset (i.e. December). Mr Williams forwarded the correspondence and asked Mr Bayfield to explain why there was not an approach to the BNP UK board until January 2006. Mr Williams then emailed Mr Bayfield stating:

“You had better call him to discuss. An e-mail trail is not I good idea. [sic]”

[247]The letter of wishes was requested by the directors of BNP UK. They wanted to have this before they were prepared to agree to enter into the put option. It provided as follows:

“... it would be advantageous for the [BNP] group if you were to grant a put option over the issued share capital of [HIL]... Accordingly, in order to assist the [BNP] group’s position, it is our wish that you grant the option under the terms of which you agree to buy the shares upon exercise by [BNP Lux]. We note that the strike price under the option is likely to exceed the market value of the shares. However, we also note that the aggregate subscription price that [BNP] will pay under the Conditional Subscription Agreement for ordinary shares in [BNP UK], will be equal to the strike price under the option. Although a loss will be incurred by [BNP UK] under the option we acknowledge that this will ultimately be borne by the [BNP] group.”

[248]It was put to Mr Peters that the concern arose as this was the proposed implementation of the plan that BNP Lux would be bought out by BNP UK at an overvalue. Mr Peters agreed that this was “so it makes a profit, absolutely. At an overvalue so it makes a profit from its involvement in the transaction.”

[249]On 30 January 2006, the board of BNP UK approved the grant of a put option to BNP Lux. It was also noted that BNP would enter into a Conditional Subscription Agreement. On 31 January 2006, the parties entered into the put option arrangements as set out above. Termination of the arrangements

[250]There is very little in the bundles in relation to the termination of the arrangements which took place on 30 November 2008. The minutes of the board meeting of HIL on 28 November 2007 record that HIL had received an Acceleration Notice dated 27 November 2007 from ALIL.

[251]There are a number of emails which indicate that Dublin branch wanted to exit from the transaction as it was having difficulties in achieving a Libor return due to the liquidity conditions in the market. Tom Woulfe of the Dublin branch sent several emails on this topic. In the first he said:

“Extraordinary liquidity conditions impacting Sumatra and ALM now claims to be losing 16 to 17 basis points on each of these deals, approximately £47,300 a month. We, SCM... have the ability to terminate both transactions on 10 business days’ notice. ALM Dublin are trying to wriggle out of this agreement. As we would like and need ALM’s cooperation for future business, I suggest we meet John halfway and offer him a monthly fee for as long as these difficult conditions continue.”

[252]However, Mr Peters’ evidence was that this was not the cause of the early termination. He said he thought that the transaction was brought to an end by A&L because they wanted to deploy the cash elsewhere, he thought as regards another transaction with BNP.

[253]The consequences of an acceleration notice were dealt with by the swap arrangements which effectively permitted BNP and A&L to cancel the swap in place between the two groups with little or no breakage costs. Further evidence on commercial and economic effects of the transaction Transaction from perspective of the London branch

[254]In his witness statement Mr Peters described the transaction as giving the London branch the opportunity to buy and sell a series of fixed and determinable cash flows over a short period of time giving rise to an attractive pre-tax rate of return (after all costs) for an overall low level of risk. He also said in his statement that “the arrangements were also attractive to the London branch as they gave a significant post tax enhancement”. From his oral evidence and the correspondence set out above, it is clear that, for the OF London team and the BNP group, the “post-tax enhancement” was rather more than a secondary attraction as this implies.

[255]Both Mr Peters and Mr Demon gave evidence in their witness statements that the dividend rights were akin to those on a bond, being a “plain vanilla” bond, as Mr Peters put it, or a “synthetic fixed income bond”, as Mr Demon put it. The fixed rate dividend equated to fixed interest on a bond and the termination dividend to the repayment of the principal amount. Both expressed the view that for the London branch this was the purchase and sale of financial cash flows of the type the branch engages in on a daily basis in the course of its banking trade. Mr Demon noted that the transaction was structured with a view to satisfying the requirements of ALIL and the BNP group but such bespoke planning was a normal part of the group’s capital markets activities.

[256]In their view, the on-going transaction with ALIL was economically equivalent to a loan at a fixed rate of interest the principal of which was repayable via the termination dividend. Mr Stanton did not disagree with this.

[257]It was put to Mr Peters and Mr Demon that the characteristics of the “loan” were unusual in that, for example, there was no debt until termination and it had features, such as that it was terminable at will, which are not typical of a bond. Mr Peters said that he had not come across a bank buying and selling an equity instrument of this type but that it was, in economic terms, akin to a bond which banks trade in daily. Mr Stanton also thought that this type of cash flow was not one which would be readily available in the market; it was created under a bespoke structure. Mr Demon accepted that having a termination provision was unusual in a bond but maintained his stance that it is not so unusual in a loan. However, he did accept that it was unusual to pay the other side’s legal costs.

[258]It was put to Mr Peters that, in fact, the BNP group had created a completely artificial financial instrument to generate something which it could buy and sell for the purposes of bringing itself within s 730. Mr Peters said:

“Well, we set up the structure, we put the structure in place. HIL was incorporated with a view to selling the strip to London and then on to [A&L] … it’s not wholly artificial. There was real commercial - I keep coming back to - this was a fundraising, this made money for the bank before any tax was taken into account … It just happens that money made was not as big or not as large as the post-tax benefits.”

[259]He was questioned on his comments in his witness statement that BNP Lux could not have sold the dividend strip direct to ALIL. It was put to him that was never contemplated because the London branch would not have been able to get the s 730 benefit. He agreed that was the case:

“No, it wouldn’t have achieved the 730. We wouldn’t have been able to achieve the 730 reduction of course if [BNP Lux] sold it directly.”

As set out above, he agreed that, absent the tax benefits, BNP Lux could have sold the dividend strip to ALIL with a parent guarantee and that, structured differently, any entity in the BNP group could have sold the strip.

[260]We accept that, looked at in isolation, the BNP group had created a product in the form of a transferable series of cash flows which, although it had some unusual features and was a bespoke product created by BNP specifically for this transaction, had the essential characteristics of cash flows that the London branch typically bought and sold at the relevant time. We accept that the fact that the product was created by the BNP group to their required specifications does not of itself preclude the purchase and sale of the product from being regarded as akin to the transactions typically carried out by the London branch. The question remains, however, whether the way the product was created and the surrounding arrangements affect whether this transaction can be said to have been undertaken in the course of the trade of the London branch.

[261]Both Mr Peters and Mr Demon sought to emphasise that, on a group basis, the arrangements in effect provided funding at an attractive rate, lower than the market rate, giving rise to a pre-tax profit on an overall basis. Both also maintained to some extent that, whilst obtaining the s 730 advantage was an important part of the transaction, the BNP group also had the purpose of obtaining this “attractive” financing. Mr Demon said in his witness statement that the transaction was ultimately approved by the TCC because it provided this funding at a low rate. Pricing of the sale of the dividend strip to ALIL

[262]We note that, as set out above, the size of the transaction was essentially determined by reference to the tax capacity available to the London branch. A&L clearly had an appetite for a higher investment. They expressed an interest in investing up to £450 million but the TCC determined the transaction size as £150 million. We accept that the agreement that ALIL would pay £150 million for the dividend strip essentially in return for a fixed rate return of 4.35% per annum, can be said to be an arm’s length transaction. There is little evidence on the negotiation between the London branch and A&L but it is reasonable to suppose that the rate would have been reached by negotiation given that A&L is a wholly independent party. Pricing of the purchase of the dividend strip from BNP Lux

[263]Having determined the transaction size as £150 million, it appears that the other pricing and profit levels were worked out backwards, in a type of reverse financial engineering, from that fixed amount to give what the London branch thought was an acceptable level of pre-tax return for each entity involved in the transaction whilst ensuring ALIL achieved its expected rate of return. None of the pricing elements of the transaction involved any actual valuation of the relevant assets in question as such.

[264]We note that it is not disputed that the amount of pre-tax profit to be realised by the London branch and by BNP Lux (through the sale of the “rump”) was determined using a “goal seek” technique to give the desired result. Whilst a discounted cash flow method of pricing the sale of the dividend strip by BNP Lux was used earlier on, this was dropped in favour of the goal seek technique. Mr Peters did not know why the methodology had changed. He had calculated a price for the purchase of the right to the dividends using the same methodology as was used before the change in approach, which gave a price of £148,383,818. Mr Stanton also noted that the price paid by the London branch for the right to the dividends was not far away from that which would have resulted from a discounted cash flow calculation.

[265]Mr Peters accepted that, on the basis that the price to be paid by ALIL was fixed, the price to be charged to BNP Lux had to be lower to give the London branch its desired rate of return. In effect, therefore, it appears that the OF team decided that £500,000 was an acceptable level of profit for the purchase and sale of this type of cash flow and plugged that into the computation of the pricing of the purchase of the dividend strip from BNP Lux to give it the desired result.

[266]Neither Mr Peters nor Mr Demon were directly involved in the pricing of the various elements of the transaction and neither had direct experience of capital markets trading. Mr Demon noted that whilst he was not directly involved in pricing he had overall responsibility for making sure that all aspects of the transaction were appropriate in his role as head of GF and he also had an overview of the group’s bond trading activities at the time. When questioned, he could not remember what he had meant by the word “appropriate”.

[267]Mr Peters gave evidence that the expected profit fees for the London branch of £500,000 (which was, in fact, £402,372) was an attractive profit for this type of transaction. It represented a spread of 33 basis points or a gross profit margin of 0.33% on the purchase and sale of the dividend rights. Banks often engaged in the purchase and sale of cash flows where the spread was low compared to the cash flow values and where in many cases only a few basis points would be earned. Mr Demon noted that the group did not have any senior bonds in the market at the end of 2005 but it did have a subordinated bond that was trading at around to 16 basis points (bid/offer spread). He said that in his experience a trader who had the opportunity to make this kind of return on buying and immediately selling cash flows would seek to take it. In his view the level of financial risk the trader was exposed to was entirely commensurate with this level of profit. Mr Demon also thought that it was to be expected that the London branch would make a profit given its role. He thought it was “entirely logical and commercial that the London branch would pay BNP Lux a lower price than it received from ALIL”.

[268]We accept the evidence of Mr Peters and Mr Demon (which was not challenged) that the level of profit realised was not out of kilter with what a trader may expect to receive for a purchase and sale of such cash flows.

[269]As regards the pricing of the purchase of the dividend strip from BNP Lux, Mr Peters provided the Luxembourg team with indicative pricing of each leg of the transaction on a number of occasions. It is clear that the Luxembourg team were aware that the price proposed for the sale of the dividend rights by BNP Lux was lower than the onward sale price to ALIL. Mr Peters said in his witness statement that the difference in the two prices reflected the fact that the London branch expected to be compensated for its role in putting the transaction together and that the branch would expect to pay less for the rights than ALIL, as ALIL would not be subject to tax on the fixed rate dividends, whereas the branch would have been taxable on those receipts rendering the dividend rate unattractively low. We note that the email of Mr Robinson dated 26 October 2005 and Mr Peters’ oral evidence confirmed that the transaction was structured to provide this as a commercial rationale for the difference in the pricing of the dividend purchase and onward sale thereby justifying the profit regarded as essential for this to be viewed as a trading transaction.

[270]As regards the commercial justification for the difference in the prices charged for the sale of the dividend strip, Mr Stanton commented in his report that, given that it was clear that the London branch was not going to hold on to the dividend strip, it was never actually going to receive the fixed rate dividends. He noted that the differing credit positions of BNP and BNP Lux may justify a different pricing but that was highly theoretical. If matters were arranged differently, such as by BNP providing BNP Lux with a parent guarantee, for example, it may have been possible for BNP Lux to have achieved the same price on a sale of the dividend strip direct to ALIL.

[271]We accept that it is clear that the London branch team were the originators and driving force behind the transaction such that it is not unreasonable to conclude that they would expect to receive a profit from it. It also makes sense that an entity which would not be taxable on the dividend receipts would be likely to pay more for the dividend strip than one which was taxable on such receipts. We note, however, that no evidence was presented that the desire to earn a profit or the difference in tax treatment were factors which were actually taken into account in determining the price. All the indications are that the figure for the desired profit was simply chosen, as the amount needed to “commercially incentivise” the London branch to enter into the purchase and sale transaction. Indeed, we do not understand the appellant to be saying anything other than that the transaction was structured so that there was a “commercial incentive” in the form of the price differential; there is no suggestion that these factors were actually considered by the London branch. As Mr Peters said, without such commercial justification, these things don’t get done.

[272]It was put to Mr Peters that, when the prices for the dividend strip and the buyout of the “rump” were being fixed, there was no negotiation between OF, London and Luxembourg based on the value of what was being bought and sold and it was essentially a cost plus calculation. He said that, whilst he was not a party to all pricing discussions, he believed that the pricing discussions were centred around the commercial profit that BNP Lux would earn from its involvement in the arrangements, “so I think that’s a fair assumption, it was a cost plus calculation is a fair assumption”. This accords with the correspondence where it is clear that Ms Majchrzak Gilot wished to obtain a particular overall level of profit for BNP Lux from its involvement (see [98(2)] above) and the final financial models which, as noted, were produced essentially using the “goal seek” method.

[273]We note that Mr Demon agreed that it may have been the case that BNP Lux received a greater profit than the London branch because the branch had the potential advantage of the s 730 benefit. Our conclusion is that all the evidence shows that the OF London team were indifferent to the level of profit for the London branch other than ensuring that it was sufficient as regards what they considered necessary for this to be a trading transaction such that the s 730 benefit could be obtained. Similarly, they appeared from the correspondence to be indifferent to the precise level of the profit in BNP Lux except as regards ensuring it received what it needed for it to be sufficiently commercially incentivised to enter into the transaction, again with a view to obtaining the s 730 benefit. On the other hand, Ms Majchrzak Gilot of BNP Lux just seemed to want a particular amount of profit for BNP Lux as indicated by her email insisting on receiving £500,000 which she said had been proposed originally (see [98(2)] above). It is not explained in any of the correspondence why this figure rose to £750,000 in the end but it seems most probable from Ms Majchrzak Gilot’s prior emails that this was at the insistence of BNP Lux. In his witness statement, Mr Peters said he recalled that Ms Majchrzak Gilot was very focussed on the quantum of return for BNP Lux and that, in his experience of having worked with her for more than 10 years when involved in transactions across multiple jurisdictions, Ms Majchrzak Gilot is very keen that the returns for BNP Lux are maximised. It does not appear that Ms Majchrzak Gilot or any of the other Luxembourg personnel were concerned with the precise level of the price BNP Lux was to receive for the sale of the right to the dividends as such except that Ms Majchrzak Gilot raised that it ought to reflect an arm’s length amount.

[274]Mr Stanton said in his report that the price received by BNP Lux was not an “arm’s length” one. In his view, the transaction was inherently loss making for BNP Lux due to it having on-going funding costs with no prospect of dividend receipts during the three year term of the transaction with ALIL. That loss was transferred to BNP in effect when it took on the funding cost by funding the purchase by BNP UK of the “rump” at an overvalue of £62.7 million. Given the loss-making nature of the transaction, Mr Stanton thought that a bank would not approve such a transaction in the absence of the prospect of obtaining the s 730 benefit.

[275]Mr Stanton was cross examined on this at some length, in particular, as regards his view that the price for the sale of the dividend strip to the London branch was not an “arm’s length” price. He accepted that he had been somewhat loose in his use of the term “arm’s length” and perhaps it was more accurate to refer to the commerciality of the pricing.

[276]He noted that it was not commercially viable for BNP Lux to sell the dividend strip at a price which gave rise to a potential loss which, had the transaction run its course of three years, would have been in the region of £9 million on a pre-tax basis. The loss was essentially the on-going funding cost to BNP Lux on the balance of its funding of around £60.6 million, assuming it had used the sale price received on the sale of the dividend rights to repay its initial investment of £210 million to that extent and that it paid interest on the balance at an assumed interest rate of 4.74%. The loss would be less on an after-tax basis assuming BNP Lux was entitled to tax relief for the interest expense (and he noted that the financial models produced for the transaction showed BNP Lux as receiving tax relief for interest). For the purpose of illustration, on an after-tax basis the loss would be around £6.4 million at an assumed rate of tax of 30%. He said that another way of putting this was that the loss arose because HIL was liable to pay tax on its interest receipts with no corresponding tax deduction for the fixed rate dividends paid to ALIL so that it had no profits to pay to BNP Lux as its shareholder.

[277]Mr Stanton calculated that for BNP Lux to make a profit or at least break even, it would need to receive, looking at this on a pre-tax basis, at least £157 million for the sale of the dividend strip or, on an after-tax basis, in the region of £154.9 million. However, in that case, the London branch would make a loss (before transactions costs) of £7 million or £4.9 million. It would not, therefore, have been economic for London branch to enter into the transaction on that basis. In Mr Stanton’s view, although the pricing of the purchase and sale of the dividend strip made sense for the London branch and ALIL, it did not make sense for BNP Lux.

[278]He noted that there were a number of problems, in his view, with pricing this transaction. The dividend cash flows are not readily available in the market. They had to be created, such that (absent this transaction) there is not a readily comparable market or a comparable uncontrolled transaction. He also noted that the cash flows were different to those on a typical bond in that either party could terminate the arrangements on short notice. He thought that a discounted cash flow basis would produce a price which was not dissimilar to that in question (as indeed Mr Peters’ calculation shows). However, the oddity was that that price would not be attractive for BNP Lux, as the seller, because of the inherent loss making nature of the way the arrangements had been set up and the dividend rights created. BNP Lux was left with an on-going funding cost on £60 million having sold all the income stream from the asset.

[279]Mr Boulton, for the appellant, referred to Mr Stanton’s report where he had said essentially that project Sumatra was a combined transaction for banking purposes and questioned why he then thought it appropriate to look at this as a single leg, in effect, assuming that what was planned (namely the sale of the shares) would not happen. Mr Stanton said that he thought this was a good way to look at it and you could then see that the same loss pops up later down the track when the parent, BNP, funded BNP UK to buy the shares in HIL from BNP Lux. Mr Stanton’s point was that essentially the loss was transferred to BNP and BNP UK as a result of BNP UK purchasing the shares at an overvalue funded by BNP. BNP UK had an impairment loss of £2.7 million and BNP had an on-going funding cost on £62.7 million.

[280]Mr Peters was questioned about this in cross examination. It was noted that the financial model which Mr Bayfield sent to Ms Majchrzak Gilot on 12 December 2005 showed that following the sale of the strip, as at 31 December 2005, BNP Lux would have a loss of £1,184,000. It was noted that Mr Stanton had calculated that BNP Lux would need to receive a price of £157 million to make a profit over the three year term of the deal with ALIL.

[281]It was put to Mr Peters that, essentially, BNP was subsidising the payment of the lower price BNP Lux actually received for the right to the dividends, given that it then funded BNP UK to purchase the shares at an overvalue to cover BNP Lux’s funding and other costs and to enable BNP Lux to achieve a profit. In other words, the effect of this was that BNP was subsidising the purchase of the right to the dividends by the London branch from BNP Lux at an undervalue. It was because BNP Lux sold at that lower price that the London branch was able to sell at £150 million to ALIL, and so, indirectly, BNP subsidised the sale of the strip to A&L. Mr Peters said:

“But when you use the term “subsidised”, I think you imply that the value of the strip sold is precisely that, knocked down, and that’s not the case. We think that the value of the strip as bought is an arm’s length price. It’s not an undervalue price.”

[282]Mr Peters noted that commercially a price of £157 million would not have been achievable. He noted that the investor or the borrower could call for the “loan” to be repaid on a three day “hair trigger”. He said: “I imagine it unlikely that someone would pay £157 million for an asset that the borrower could repay three days later at £150 million ….and [ALIL] of course would not have paid that”.

[283]He continued to note that the borrowing rate in BNP Lux, at 4.74%, was higher than the deposit rate in HIL, of 4.59%, which meant that BNP Lux would have a high interest cost whatever price it had sold the right to the dividends at:

“So on a pre-tax basis, BNP Lux would have high interest costs, whatever the number is based upon … and until such time as dividends flowed up to Luxembourg, it would have losses”. He concluded from this that: “[BNP Lux] wanted to make sure that both ends happened. I don’t believe it’s a case of [BNP Lux] not selling the strip were it not be able to sell the rump. I think [BNP Lux] would not have entered into the transaction at all and capitalised HIL.”

[284]He accepted that BNP Lux “absolutely would make a loss unless it was able to exit the transaction by selling the rump”. He accepted that ALIL would not have paid £157 million for the strip, and, BNP Lux would not have done the transaction, without being assured of being bought out at a price which covered its costs and gave it a profit and, BNP UK would not have granted the put option without the promise by BNP to subscribe an amount of share capital equal to the price. He said:

“The point is that each entity involved in this series of transactions was commercially incentivised to do so. So from an individual company perspective and a group perspective, this made commercial sense and this is before - the pre-tax level before we talk about 730…. If the transaction hadn’t made commercial sense to each entity, it would not have happened.”

[285]It was put to him that the problem about this whole structure is that the dividend rights being sold were not worth the cost incurred in creating them. Mr Peters said:

“the cost that you are adding to the £150 to get to £157 is a funding cost, an interest cost. It’s not to do with the value of the shares or the value of the strip. If we look at sort of the commerciality and say, you know, looking at cash flow, what money went in and what money went out … £210 goes out from [BNP Lux] on day 1 when it subscribed for the shares. £149.1 then comes in a day later. There’s some funding costs and then £62.7 comes in at the end to give rise to the overall profit for [BNP Lux] … [whilst it was covered by BNP] But it is not given as a subsidy for a strip price, it is given to ensure that some of the cash flows - the transactions that [BNP Lux] is being invited to enter into result in [BNP Lux] making a profit……. So what we’re saying is the sum of the rump and the strip is not equal to the sum of the whole, what was contributed on day 1.”

[286]He agreed that, looking at the transaction at the point of sale of the dividend strip, BNP Lux had an asset which was worth less than £60 million. He noted that BNP Lux had the right to terminate the transaction on short notice on HIL paying the termination dividend but acknowledged that was not the intention given the relationship with A&L. We note that Mr Peters produced some calculations in his witness statement in support of the fact that the price paid by ALIL and that paid to BNP Lux reflected an arm’s length rate. We have set this out in the section below regarding the effect of the “loan” from ALIL as the calculations are based around that analysis.

[287]Mr Peters disputed Mr Stanton’s view of the effect of the sale of the “rump” at an overvalue. He said that, contrary to Mr Stanton’s view, the loss was not in effect transferred to BNP when it funded the acquisition of the “rump” at an overvalue thereby taking on a funding cost to the tune of £62.7 million. He said this was because the amount paid by BNP merely moved funds from one group subsidiary (BNP UK) to another (BNP Lux) and overall, looking at the cashflow schedule he produced, the group made a profit on a pre-tax basis. For example, he said: “[BNP Lux] has sold an asset at that point arguably at £2.7 million overvalue, and so therefore has a gain. [BNP UK] has purchased an asset at £2.7 million overvalue and therefore has a £2.7 million loss. [BNP Lux] and [BNP UK] are subsidiaries of [BNP], and therefore the effect on [BNP] and its shareholdings is nil as a result of doing that. But if you look at just one leg, I accept that you have a loss to BNP on its shareholding of BNP UK…... If you look at BNP as an entity and look at all the other entities, and for this purpose let’s assume London branch and Dublin branch are separate legal entities...........each of those entities makes a pre-tax commercial profit from this arrangement, and when you add those entities together, that pre-tax commercial profit is greater than the pre-tax expense at the level of BNP. BNP has effectively done two things: it has funded £60 million, or £62.7 million to be precise, £62.7 million in [BNP UK], and £59 million of that it has funded at a fixed rate because it has entered into an interest rate swap fixed for floating against Dublin branch…… and £3.5 million we have assumed floating because there's no swap in place, and this capitalisation of [BNP UK]”.

[288]Similarly, he noted that any parent company that raises debt to equity fund a subsidiary is going to make a loss until such time as the dividends come up from the subsidiaries. He again referred to treating the Paris, Dublin and London branches as separate legal entities. He noted that from a BNP management perspective, they have their own management boards, territory management and approvals systems in each place. He said:

“Until such time as the income made by the subsidiaries is distributed up to the parent, the parent is always going to have a loss. If a parent chooses to put some capital into - as parent puts an asset cash into subsidiary A, and subsidiary A uses that to acquire something from subsidiary B at overvalue, you’ve just got a profit and a loss in those two entities that net off.”

[289]We accept that, looking at the price achieved on the onward sale to ALIL, the price paid by the London branch to BNP Lux can be said to be an “arm’s length” price, in the way in which that term is commonly understood, as the price which could be expected to be paid/received between unconnected parties. Describing the price as falling outside that parameter, as it is typically applied, is not a correct reflection of the transaction. We have found that the transaction made with ALIL can be said to be made on “arm’s length” terms and that the transaction size of £150 million was the starting point for the pricing of the elements of the transaction. On that basis, it can hardly be said that a price of £157 million or anything above £150 million would be an “arm’s length” price. In creating a right to a termination dividend of £150 million and to a monthly fixed rate dividend at 4.35% per annum, the London branch had effectively set the parameters for what a third party could be expected to pay to acquire those rights.

[290]However, it is clear from the economics that the underlying point made by Mr Stanton is correct and is undisputed; the transaction was loss making from BNP Lux’s perspective absent the sale of the shares. This was due to its on-going funding cost on around £60 million of the outstanding funding it had used to invest in the shares in HIL, in circumstances where it was not going to receive any income from those shares over the scheduled three year period of the transaction with ALIL. The excess of £60 million of funding had to be put into HIL in order to enable it to pay its corporation tax liability on the income it received on its deposit given that it was not able to obtain a tax deduction for the fixed rate dividends paid to ALIL out of that income. Another way of looking at this, therefore, is that the loss represents the cost of obtaining a loan of £150 million with no tax deduction for the fixed return paid.

[291]This loss or cost is recognised in the papers presented by GF and OF to the TCC and in the later GF memo circulated on 10 January 2006 (see [229] above). Mr Demon agreed that the loss of over £6 million identified in that memo was essentially a funding cost on £60 million or that it arose because the £150 million of funding “was not tax deductible” which he said was the same thing. Both Mr Peters and Mr Demon were clear that the transaction was uneconomic but for the availability of the s 730 benefit. Mr Demon said he would not have approved the transaction if the London branch did not have the required tax capacity to utilise the s 730 benefit and unless he thought that benefit would be obtained as otherwise the group would have a net loss. The economic viability of the transaction was, therefore, accepted to be entirely dependent on the obtaining and utilisation of the s 730 benefit.

[292]Both Mr Peters and Mr Demon maintained, however, that this was just a very tax efficient way of obtaining the “loan” of £150 million from ALIL. They both said that the non-deductibility of the fixed rate dividends should not be looked at in isolation. In effect, that disadvantage was more than compensated for by the s 730 benefit. Mr Peters said that the position was not much different to a plain vanilla loan where the borrower would want to get a tax deduction for the interest paid; it was just that, in this case, the deduction more than covered the interest cost by some considerable way. Mr Demon said that this was a “special way of borrowing” where there was no deduction for the “interest” but a deduction “on top for the principal of the loan”. He said that the commercial purpose was to borrow and the transaction allowed this to be done at a post-tax attractive interest rate or quasi-interest (in other words, taking into account the s 730 benefit). We comment on this in the discussion on the trading issue below.

[293]From BNP Lux’s perspective, Mr Peters was clear that it would not have entered into the transaction at all if it were not assured that it would be able to exit on recovering its funding costs and realising a profit. That the intention was that this would be done through the sale of the shares to a UK company at an overvalue was evident from the early stages of the planning. The BNP Lux team were clearly focussed on the potential exit and, as noted, were insistent upon making a profit of a specified amount. The London OF team delayed putting anything formal in place, despite the obvious pressure from BNP Lux to do so, because of concerns raised by the UK tax team that concluding a sale in the same financial year as the sale of the dividend strip took place would prejudice the tax planning. It is clear from the correspondence that it was a UK tax concern (see [216] above). It appears that some of the presentation in the documents, therefore, downplays the likelihood of this happening but it is clear that this was planned.

[294]We note that the TCC made a recommendation that other possible purchasers of the shares should be considered. The early version of the TAC minutes recorded that it was proposed that possible purchasers outside the group should be considered and, the later version of those minutes, that other group purchasers should be considered. However, as accords with the correspondence on this topic, Mr Peters confirmed that there was never any real attempt to source any alternative purchaser. Indeed, that anyone other than a group company would buy the shares at an overvalue appears wholly unrealistic. We note Mr Demon was reluctant to accept that alternatives were not considered and that, in reality, the sale of the shares to BNP UK was practically certain, at any rate by early January. However, in the light of the correspondence on this topic, we find his view is unrealistic.

[295]In any event, Mr Peters was very clear and so is the correspondence that, whatever structure was adopted to achieve this, BNP Lux was expecting to be bought out of the structure at a profit. It would not have provided the financing required to create the dividend strip if this were not the case. Somewhat ironically, given the desire on the part of the UK tax team to avoid any firm plan for the sale of the shares at the time in question, this is now put forward by the appellant as a positive feature. In their view, the two steps, namely the sale of the dividend strip by BNP Lux for a price which left it in a loss making position and the sale of the “rump” at an overvalue, must be looked at together. Looking at both limbs and, as they now acknowledge, given that the sale of the “rump” was always going to happen, BNP Lux was suitably “commercially incentivised” to enter into the transaction.

[296]We also consider that it is clear that the loss arising from the on-going funding cost was, in effect, transferred to BNP UK and BNP on the acquisition of the “rump” at an overvalue, thereby leading to a loss for BNP UK of £2.7 million and an on-going funding cost for BNP on £62.7 million. Mr Peters thought that this did not render the transaction uncommercial for the reasons set out above. In particular, he noted that notwithstanding the funding costs, overall the group made a before-tax profit of £1.13 million from the transaction. We note, however, that BNP fully accepted that this profit would be wiped out by the tax cost in the form of the non-deductible fixed rate dividends but for the availability of the s 730 benefit. We comment on this further in our conclusions below. Effect of “loan” from ALIL

[297]It was not disputed that the effective rate of borrowing on the “loan” made by ALIL to the BNP group was at a discount to Libor such that a small profit was generated to the group overall on a before-tax basis. Mr Peters said that the reason why the effective dividend rate was relatively low was because ALIL was not taxable on the dividends. This meant that it was prepared to accept a rate lower than prevailing interest rates given that it would have been taxable on such interest receipts. He asserted that, in effect, part of the benefit of ALIL receiving a return which was not taxable was shared with the BNP group; it was the sharing of that benefit which enabled the BNP group to earn its pre-tax profit. He thought that the calculation of the dividend rate was based on a commercially agreed split of this benefit as to 80% for ALIL and 20% for the BNP group. He referred to Mr Williams’ email of 13 December 2005 in support of this (see [221] above). Mr Demon essentially agreed with this economic analysis.

[298]Mr Peters tried to illustrate this by calculating the benefit or cost to the parties, assuming the deal ran its full three year term, so that the total amount of dividends paid would be £19,590,792.25 and the notional interest expense due from ALIL would be £21,349,479 (calculated at the swap fixed rate of 4.74%). On that basis, the difference between the total interest payments/receipts and the total dividends would generate a pre-tax loss for ALIL and a corresponding profit for BNP of £1,758,687.20.

[299]On an after-tax basis, however, ALIL would have an overall return of £4,646,156.64. This was the case assuming it obtained a tax deduction for the notional interest cost and that the dividends were not taxable giving a tax saving of £6,404,843.84 (assuming the applicable corporation tax rate was 30%).

[300]Mr Peters calculated that the overall after-tax positive receipt for ALIL of £4,646,156.64 was the equivalent of a receipt of £6,637,366.62 if ALIL had instead received taxable interest income. (This figure is in effect £4,646,156.64 grossed up by the corporation tax rate of 30%). He and Mr Stanton referred to this as the pre-tax equivalent.

[301]He concluded that the 80/20% split of the benefit to ALIL of receiving non-taxable dividend income was represented by £6,637,366.62 for ALIL (being 80%) and £1,664,687.20 for BNP (being 20%).

[302]In his oral evidence, Mr Peters said that he thought that a better way of looking at this was to look at the comparable dividend receipts and interest receipts only. On that basis, ALIL received a return on an after-tax basis of the full amount of the total dividends of £19.5 million as the dividends were not taxable. If it had made a loan in return for taxable interest income of £21.3 million, on an after-tax basis, it would have received a return of around £15 million only. Therefore although, in pure income terms, ALIL received less on the dividend transaction than on an interest transaction, on an after-tax basis it was better off due to the tax saving.

[303]Mr Stanton did not have any issue with Mr Peters’ calculation except that he thought, as explained below, that Mr Peters was not comparing like with like for A&L and the BNP group as he did not set out the after-tax position for the BNP group. He noted that the way banks evaluate is based on the way they operate, namely that the bank has a large Treasury department which raises funds for the bank which lends front-line operations departments at Libor and then lends out at a margin over Libor. Thus, all their performance and measurement is based on an internal rate of borrowing. He noted that, according to its accounts, ALIL did not have an interest cost because, for tax purposes, it is an investment company, “and it’s quite good for investment companies just to have share capital coming in and investments coming out”. However, he thought it likely that somewhere along the line, the business department within A&L that set the transaction up, would be charged interest on the funds which in effect were used to acquire the dividend rights.

[304]In his witness statement, Mr Peters likened the transaction with ALIL to a preference share lending or deposit surrogate transaction. He said that, at the time, such transactions were common and a split in the ratio of 70/80% for the investor and 20/30% for the borrower was typically that used for sharing the tax benefit of those transactions. A deposit surrogate involved a bank issuing preference shares to corporate or institutional investors on which a fixed rate dividend was payable at a rate lower than the market rate of interest. As the dividend was not taxable for the investor, it received a higher post tax rate of return than it would have done on an equivalent borrowing at interest. The bank would, therefore, in effect borrow at a lower rate of interest than the prevailing rate of interest thereby achieving a pre-tax benefit. The bank would typically be indifferent to the lack of a tax deduction for the dividend (if it had excess losses such that the deduction was of no benefit to it) or would in some way achieve the economic equivalent of a deduction (such as through a repo transaction). These would be priced typically by comparing the pre-tax equivalent return expected for the investor (being the amount the investor would have to receive to generate the expected return if it were taxable on the income) with the pre-tax return for the borrower. The actual dividend rate would be adjusted/set to achieve the desired ratio of return commercially agreed between the parties. His understanding at the time was that A&L viewed the transaction as akin to that type of transaction and had agreed the pricing on a similar basis.

[305]Mr Stanton thought that the transaction was not analogous to a deposit surrogate because, essentially, HIL was a taxpaying entity which was not indifferent to the lack of availability of a tax deduction (as it did not have surplus losses making such a deduction useless to it) and was not going to obtain the economic equivalent. In his view the situation had a greater similarity to a preference share lending although he thought that technique did not really work here. He stated that preference share lending is a technique which works well for a non-taxable borrower (which is not affected by the lack of a tax deduction for the dividend) and a UK tax paying entity (which does not pay tax on the receipt). In that case, the rate of return on the “lending” is fixed so that essentially the parties share the benefit of the fact that the investor is not taxable on the return. He said, however, that was not the situation here because HIL was a taxpaying entity so that it was affected by the lack of a tax deduction for the fixed rate dividends. Hence the on-going cost of the transaction on an after-tax basis, absent the s 730 benefit.

[306]Mr Stanton stated that Mr Peters was not comparing like with like in his calculation. Whilst he looked at ALIL’s after-tax position, in assessing the “benefit” to the BNP group, he looked only at the position before the tax effect of the transaction was taken into account. Taking into account the fact that HIL would not get a tax deduction for dividends paid would give an overall after-tax loss (equal to ALIL’s after-tax benefit) of £6,404,843.84. In his view, there was no “benefit” which could be shared in such circumstances (at any rate disregarding the s 730 benefit). As set out in [44] above, Mr Stanton’s view, which was not challenged, was that, in the absence of the s 730 benefit, the rate of “interest” on the funding from ALIL was actually very expensive and well above the BNP group’s usual cost of funding at the time.

[307]It was put to Mr Peters that, in such circumstances, it can only be said that the lender receives a benefit on receiving a return in a non-taxable/non-deductible form if the borrower is not a tax paying entity. He agreed that unless the bank was indifferent to the lack of a tax deduction because it had losses or the group obtained a deduction at least equal to the economic cost of the dividends paid, “you are absolutely right, it’s unlikely that this would be considered economically attractive on a post-tax basis, because there wouldn’t be sufficient benefit from the investor’s end to bring you both to a profitable position.”

[308]He acknowledged that his figures did not look at the post-tax cost to HIL. He was asked whether it would have been more logical to look at the reality of a company which, as HIL did, had an immediate tax exposure. He said:

“Yes, but in the structure that was executed, we achieved the deduction at least equal to that exposure in the London branch. So the benefit that is being shared here is [A&L’s] benefit, and [A&L] would have been indifferent as to whether we suffered tax at 10%, 50% or no tax at all … But the tax cost it [HIL] suffers is one part of the BNP group’s - in this case in the UK - overall tax. So if – let’s imagine that you are tax paying as a group. Does it matter from a group perspective if it’s paid out of one subsidiary or another? The group as a whole is tax paying … It [HIL] may be de-grouped for tax purposes, but the economics of this entity flow through fully to BNP.”

[309]It was put to Mr Peters that the tax benefit was achieved upfront and the longer the transaction went on, the more that benefit eroded due to the tax cost involved in paying non-deductible fixed rate dividends. He acknowledged that there was a tax cost and, similarly as for a preference share lending style transaction, “you needed somewhere in your group to have a deduction at least equal to the cost of the dividends that were paid to your third-party investor. This transaction, it was just significantly higher. So, the longer the transaction went on for the longer there would be dividends paid for which there would be no tax deduction.”

[310]We note that the papers prepared by GF, OF and UK tax for the purposes of the approvals process all acknowledge that the tax benefit was obtained upfront. The statements in the GF and OF papers suggest that the subsequent cost of the non-deductible fixed rate dividends was seen as eroding this benefit. The email from Mr Bayfield on 19 November 2005 (see [106] above) was explicit that the best thing for the group was to terminate the transaction immediately after it was implemented: “the rational thing for the [BNP group] to do is to pull the transaction on day 2 because we will have secured our tax advantage, and therefore paying [dividends] to [A&L] at the rate agreed is a drag on the group”. It is clear from the correspondence and it was accepted by Mr Peters that the reason the BNP group would not terminate was because of the relationship with A&L. We also noted that matters were structured so that, if there was an early termination, ALIL would not suffer any material cost; that was the reason for the arrangement regarding ALIL’s legal costs and for the provisions in the swaps resulting in no material breakage costs.

[311]It was put to Mr Peters that it was a bit odd to say that the BNP group was raising funds cheaply but that the longer the loan went on the more expensive it became. Mr Peters said that this was not the way the team looked at it “we had the headline rate - borrowed at a cheaper rate … and at the same time achieved the deduction via 730 that was at least equal to the costs of the dividends that we paid. Therefore, we were indifferent to the fact that the dividends themselves were non-deductible”.

[312]In re-examination, Mr Stanton gave the view that the loan funding in this case was not very significant for BNP and was not very “helpful” funding because of its short-repayment terms. As set out in [154] above, Mr Peters agreed that the use of the funds was somewhat restricted given the termination rights and gave the view that, in effect, the benefit of the funding accrued on a daily basis whilst it remained in place.

[313]We find it difficult to see that in these circumstances where, absent the s 730 benefit, the funding obtained would be very expensive, it can be said that obtaining funding at a cheap rate was an objective of the transaction; as Mr Demon and Mr Peters say, the transaction was also designed to obtain the s 730 benefit but that simply meant this “cheap funding” was obtained in a highly tax efficient way. The evidence seems to us to indicate quite the contrary, namely that the objective was solely to obtain the s 730 benefit, with obtaining financing at an otherwise expensive rate being necessary in order to achieve that objective. Again, we have set out our further conclusions below.

[314]Finally, in his second witness statement, Mr Peters expanded upon the calculation above to support the view that the price paid by ALIL and, accordingly, that paid to BNP Lux for the right to the dividends reflected an arm’s length commercial price. He did this by setting out a number of scenarios assuming a different ratio for the sharing of the “benefit” to ALIL, thereby showing a higher or lower price for the dividend rights depending on the ratio, but otherwise using the same assumptions as used in his initial calculation. Not surprisingly, the higher the price, the lower the benefit to ALIL and the greater the pre-tax return to BNP (and vice versa). Mr Peters concluded from this that the minimum price which the London branch might have been prepared to sell at would have been in the region of £148.5 million (where ALIL obtain all of the benefit) and the realistic maximum price which ALIL would have paid was in the region of £150.6 million (reflecting a ratio of the sharing of the benefit of 70/30%). He noted that the price actually paid to BNP Lux fell within these parameters.

[315]Mr Stanton said that he did not agree that the pricing would be done on this basis. In particular, Mr Peters did not adjust the dividend rate but kept that constant. We note that in his witness statement (as set out above), Mr Peters referred to the actual pricing being done by adjusting the dividend rate to give the desired ratio of benefit sharing. In any event, we consider that, as Mr Peters was not involved in the actual pricing discussions, there is little weight to be attached to his financial models provided after the event and supported by only a single reference in the correspondence to pricing by reference to the sharing of ALIL’s benefit in Mr Williams’s email. We note that the email was prepared for internal purposes in relation to a DOTAS issue.

[316]We also note that, contrary to the manner in which this was presented by Mr Peters, the GF memo refers to sharing of the s 730 benefit with ALIL in the ratios of 90% for the BNP group and 10% for ALIL. This benefit sharing is also it seems reflected in the internal allocation of 90% of the discounted benefit (of 75%) to the London OF team. It does not appear from this, therefore, that BNP considered they were sharing in ALIL’s tax benefit but that they were sharing their benefit with ALIL. Trading issue Overview of the issues

[317]The first question is whether the London branch undertook the transaction as part of its established banking and financial trade, such that it was entitled, in computing the profits from that trade, to deduct the price paid for the right to the dividends. This is essentially a question of fact; there is no statutory definition of the relevant terms.

[318]Much of the case law in this area has centred on determining the line between what may be categorised as an adventure in the nature of a trade and, on the other hand, an investment or capital transaction. In that context, the often quoted “badges of trade” described by Sir Nicolas Browne-Wilkinson V-C in Marson (Inspector of Taxes) v Morton

[1986] STC 463

at 470 to 471 set out a number of indicators of a trading transaction (such as whether it was a one-off transaction, whether it is related to the trade, the nature of the subject matter, the way in which the transaction was carried through, the source of finance of the transaction, whether there was an intention to re-sell in the short term or to hold for a lengthy or indefinite period).

[319]In this case, however, there is no dispute that the London branch had an established financial trade and that, on the face of it, the transaction comprised the elements which would usually mean that it was undertaken as part of that trade, being the purchase and sale of a financial asset, of a type the London branch usually dealt in (albeit one created under bespoke arrangements) for a profit. The debate is whether and how the London branch’s undisputed aim of obtaining the s 730 benefit affects the analysis.

[320]At the heart of the issue is the effect of the House of Lords decisions in the “dividend stripping” cases of FA & AB Ltd v Lupton (Inspector of Taxes) [1971] UKHL TC 47 580 (‘Lupton’) and Thomson v Gurneville [1971] UKHL TC 47 633 (‘Gurneville’) which were heard within days of each other in the 1960s following on from two earlier conflicting decisions by the House of Lords in Griffiths v JP Harrison (Watford) Ltd

[1963] AC 1

(‘Harrison’) and Finsbury Securities Ltd V IRC

[1966] 1 WLR 1402

(‘Finsbury’). The essential elements of these cases were the same. A share dealer bought shares in a company containing distributable profits, stripped out the profit in the company by way of dividend and wrote down the value of the shares. Under the rules at the time, this enabled the dealer to claim a repayment of tax due to the reduction in value of the company notwithstanding that there was no real financial loss. As here, the dividend stripping transactions comprised the elements usually associated with a trading transaction of the type in question. It was held, however, in both Lupton and Gurneville, that the transactions were not undertaken in the course of the share dealing trade essentially as they were undertaken wholly or mainly as tax recovery devices rather than for commercial purposes of dealing in shares. They enabled the taxpayers to recover substantial amounts of tax where there was no actual financial loss.

[321]The parties’ detailed submissions on the cases are set out in the discussion section but, in summary, their views were as follows. BNP submitted that Lupton and Gurneville established that the question of whether a transaction is a trading one is to be assessed objectively and that motive (whether fiscal or otherwise) or tax effect is irrelevant unless the transaction is one which according to its features, as objectively assessed, is ambiguous or equivocal. BNP considered that, in applying this objective analysis, whilst the context of the transaction is relevant, according to later cases such as Samarkand Film Partnership No 3 v HMRC

[2015] STC 21350

(‘Samarkand’), Lupton indicates that the focus is very much on the particular transaction according to the terms on which it is made.

[322]BNP argued that, on such an objective assessment, the transaction is plainly a trading transaction; the London branch acquired the right to the dividends for onward sale to ALIL at a profit as part of its established financial trade. That the London branch wanted to obtain the s 730 benefit is irrelevant. Such a motive cannot alter the shape or nature of what is, objectively speaking, unambiguously a trading transaction. It is not a case, like Lupton, where the very terms of the transaction itself were dictated by tax considerations. The obtaining of the s 730 benefit did not intrude on the transaction at all; s 730(3) operates quite independently of any terms of the transaction.

[323]BNP submitted that, in any event, even if it could be argued that the transaction is ambiguous on an objective analysis, motive or purpose is no longer relevant in such an equivocal case either. The later cases of Ensign Tankers (Leasing) Ltd v Stokes

[1992] 1 AC 655

(‘Ensign Tankers’) and New Angel Court Ltd v Adam

[2004] 1 WLR 1988

(‘New Angel Court’) have established that the proposition that motive is relevant in equivocal cases is no longer good law

[324]HMRC did not suggest that Lupton creates some special principle of tax law (akin to how the Ramsay principle was once viewed) that disqualifies something which would otherwise constitute a trading transaction from being trading. Rather, they argued that Lupton is simply authority for the proposition that, where activities are carried out to facilitate a tax avoidance scheme or have been materially altered to accommodate a tax avoidance purpose, on a realistic analysis of all the relevant circumstances, these activities do not constitute trading even though they bear some of the outward appearances or trappings of a trading transaction.

[325]HMRC did not accept that purpose or motive has no role to play in assessing whether a transaction is undertaken in the course of the trade or that the decisions in Ensign Tankers or New Angel Court have that effect. In any event, in their view, the assessment has to be made by reference to the broader picture as to how the dividend rights were created and the nature of the on-going transaction (on the basis of Lupton, Gurneville and Samarkand). On that basis, it is clear that the transaction was unequivocally, as in Lupton and Gurneville, a device for obtaining a tax advantage. Case law

[326]In Harrison, the taxpayer bought shares in a company for £16,900 and, following receipt of a dividend of nearly £16,000, sold the shares for £1,000. The taxpayer claimed to set the resulting loss against the net dividend and that it was entitled to a repayment of tax. The majority of the House of Lords, which included Lord Morris and Lord Guest, decided that the repayment of tax claimed by the taxpayer was allowable as the transaction was carried out in the course of a trade of dealing in shares. The nature of the transaction, as a trading transaction, was not negated by the taxpayer having a fiscal motive.

[327]Lord Morris said that a trading transaction does not cease to be such because “it is entered into in the confident hope that, under an existing state of the law, some fiscal advantage will result”. He said:

“In judging as to the essential nature of a transaction it will often be relevant and of assistance to consider the objects and intentions which are the inspiration of the transaction. In the present case, however, I cannot think that there is room for doubt as to the essential nature of the transaction: it was a transaction which was demonstrably of a trading nature and it was not divested of that nature merely because it was entered into with the expectation that as a result (but not as part of the trading activity of the company as such) some tax recovery might be claimed.”

[328]He concluded that:

“The possibility of tax recovery may be a result made possible by the trading activity but I am unable to accept that if a transaction fairly judged has in reality and not fictitiously the features of an adventure in the nature of trade it must be denied any such description if those taking part in it had their eyes fixed upon some fiscal advantage.”

[329]Lord Guest also explicitly rejected the argument that, in order to ascertain whether there was trading, it was relevant to look at the object, result or intention of the activity:

“The test is an objective one. The question to be asked is not quo animo was the transaction entered into but what in fact was done by the company … I therefore conclude that neither the fact that the company intended to make a loss nor the fact that the company intended to make a fiscal advantage out of the transaction negatives trading. In my opinion one has to look at the transaction by itself irrespective of the object, irrespective of the fiscal consequences, and ask … ‘whether the operations involved in it are of the same kind, and carried on in the same way, as those which are characteristic of ordinary trading in the line of business in which the venture was made.’ The company had power to deal in shares, they bought shares, they received a dividend on these shares, they sold the shares. This was just the ordinary commercial transaction of a dealer in shares.”

[330]In Finsbury, a share dealer bought the shares in 15 companies which carried the right to special dividends of an amount equal to the company’s anticipated net profits from its business in the specified period, up to a stated maximum. The dealer bought those shares for the stated maximum (which was to be reduced to the extent that the relevant dividends fell below that maximum) plus one-half of any referable tax repayments on any loss claim it made. Each year the value of the shares fell by reason of the dividends paid and the dealer claimed repayments of tax on the resulting loss.

[331]Lord Morris (with whom the other Lords agreed) concluded that whilst the arrangements were not a sham, they were “no more than devices which were planned and contrived to effect the avowed purpose of tax avoidance”. He distinguished Harrison on the basis that the circumstances were “essentially different” as the sellers in Finsbury were involved in obtaining a share of the tax repayment. In Harrison, the transaction was plainly a trading transaction and that there were “good grounds for thinking that welcome fiscal benefit could follow did not in any way change its character”. The sellers of the shares had no further concern once they had sold; the purchaser was free to deal with them as it wished On the other hand “the essence of the arrangements” in Finsbury “was that the future interests of the vendors were being safeguarded” in that they were “to have all the benefits that would have resulted from their shareholdings had there been no scheme” whilst being “saved from the full extent of the exactions which taxation imposes” and “it was of the essence of the scheme that the company should continue to hold the shares during the periods covered by the particular sets of transactions”.

[332]He concluded that the transactions in Finsbury “were in no way characteristic of nor did they possess the ordinary features of the trade of share dealing”. The shares were not acquired for the purpose of dealing with them. In “no ordinary sense were they current assets” given they had to be retained as part of the scheme. The transaction on its particular facts was not “…‘an adventure or concern in the nature of trade’ at all. It was a wholly artificial device remote from trade to secure a tax advantage.”

[333]In Lupton, Lord Morris continued to take a similar approach. He started, at 617 I, by noting that some help may be derived from considering decisions of courts as to how other transactions have been regarded but “the question for decision will be whether the particular transaction under review can and should be regarded as a trading transaction within the course of the trade of a dealer in shares”.

[334]He stated, at 617 I to 618 A, that the nature of a transaction, which is plainly a trading transaction is not altered according to whether or not it is profitable or how it is treated for tax purposes:

“This enquiry may not involve or necessitate a consideration of the profitability of a transaction or of the tax results of a transaction. One trading transaction may result in a profit. Another may result in a loss. If each of these, fairly judged, is undoubtedly a trading transaction its nature is not altered according as to whether from a financial point of view it works out favourably or unfavourably. Nor is such a transaction altered in its nature according as to how the revenue laws determine the tax position which results from the financial position.”

[335]Lord Morris drew a distinction, at 618 C to H, between cases where a person takes advantage of the fiscal result provided for by the legislation under an “ordinary” trading transaction and more engineered “hybrid” transactions devised by specialists and sought out by the sellers of shares. He said that if any of these specialists are to be found amongst those whose ordinary trade is that of dealing in shares it must be said that “in the fashioning of these tax-engineering operations they may be stepping aside from the paths of their trade”.

[336]He again advocated, at 619 G to H, an objective approach:

“… once it is accepted, as it must be, that motive does not and cannot alter or transform the essential and factual nature of a transaction, it must follow that it is the transaction itself and its form and content which are to be examined and considered. If the motive or hope of later obtaining a tax benefit is left out of account, the purchase of shares by a dealer in shares and their later sale must unambiguously be classed as a trading transaction.”

[337]In his view, at 619 H to I, the transactions in the Harrison case were “solely and unambiguously trading transactions”:

“There was a purchase of shares and after receipt of a dividend a sale of shares. There was no term, express or implied, in any contract or any transaction which in any way introduced any fiscal element. No fiscal consideration or arrangement intruded itself in any way into any bargain that was made. There was merely an acknowledged reason which inspired one party to enter into certain trading transactions. If that party later made some tax claim that claim would be no part of a trading activity.”

[338]He continued, at 620 A, that Harrison was not a case where there was any non-trading element, seeming to suggest that if it were “equivocal” or ambiguous, motive may be relevant:

“The transactions in the Harrison case not only had all the characteristics of trading, there was no characteristic which was not trading. There was nothing equivocal. There was no problem to be solved as to what acts were done. To the question - quid actum est? there could be but one answer. The question - quo animo? was irrelevant.”

[339]He continued to note, at 620 C to D, that it had been suggested that the Harrison case decided that a transaction can be a trading transaction even though “it is a pure dividend-stripping transaction entered into with the sole object of making a fiscal profit without any view to a commercial profit”. He said that such a suggestion “is ill-founded and misleading.” He again indicated that the distinction between a trading case, such as Harrison, and a non-trading case, such as Finsbury, depended on whether another party was involved in the tax benefit:

“The word ‘transaction’generally suggests some arrangement between two or more persons … But [in Harrison] there was no arrangement whatsoever under which the sellers to Harrisons of the shares or the purchasers from them of the shares were concerned as to whether Harrisons would or would not later make some claim which under the law as it then stood they might be able to make. There was, therefore, no dividend-stripping ‘transaction’ in the Harrison case in the sense that any other person had any control or concern or interest as to what Harrisons would do once they had bought the shares.”

[340]He then, at 620 F to G, contrasted Finsbury on the other hand as being “wholly and fundamentally different” being of a nature which he described as follows: “It is manifest that some transactions may be so affected or inspired by fiscal considerations that the shape and character of the transaction is no longer that of a trading transaction. The result will be not that a trading transaction with unusual features is revealed but that there is an arrangement or scheme which cannot fairly be regarded as being a transaction in the trade of dealing in shares”.

[341]So, in his view, at 620 G to H, in Harrison there was “not a trace of any fiscal arrangement”, whereas in Finsbury “certain fiscal arrangements were inherently and structurally a part of the transactions” or “were central to and pivotal of the transactions under review”. He concluded that Finsbury was an example of a case where, as Megarry J had said in the lower court, “the fiscal element has so invaded the transaction itself that it is moulded and shaped by the fiscal elements”. He continued, at 620 I to 621 A to C, that this was helpfully expressed by Megarry J as follows ([1968] 1 W.L.R. 1401, 1419):

“If upon analysis it is found that the greater part of the transaction consists of elements for which there is some trading purpose or explanation (whether ordinary or extraordinary), then the presence of what I may call ‘fiscal elements’, inserted solely or mainly for the purpose of producing a fiscal benefit, may not suffice to deprive the transaction of its trading status. The question is whether, viewed as a whole, the transaction is one which can fairly be regarded as a trading transaction. If it is, then it will not be denatured merely because it was entered into with motives of reaping a fiscal advantage. Neither fiscal elements nor fiscal motives will prevent what in substance is a trading transaction from ranking as such. On the other hand, if the greater part of the transaction is explicable only on fiscal grounds, the mere presence of elements of trading will not suffice to translate the transaction into the realms of trading. In particular, if what is erected is predominantly an artificial structure, remote from trading and fashioned so as to secure a tax advantage, the mere presence in that structure of certain elements which by themselves could fairly be described as trading will not cast the cloak of trade over the whole structure.”

[342]Lord Morris then turned to the particular facts of the case. He placed emphasis on the fact that the sellers sold the shares to the dealer on the basis that the price to be received was dependent upon the success or failure of the tax-recovery claim which the dealer undertook to make. He thought it was apparent, therefore, at 621 G, that the transaction was not one in which any possible tax consequences resulting to a purchaser of shares would be his concern only but “it was one in which there was a carefully worked out scheme which (in the hope of mutual financial benefit) was shaped and moulded by the fiscal possibilities”.

[343]Lord Morris noted, at 622 I to 623 A, that the vendors were therefore directly and financially interested in the result of the purchaser’s promised claim and concluded as follows:

“it was submitted that the truly strange arrangements which I have summarised were but the arrangements of a trading transaction of a dealer in shares. It was further submitted that the elaborate and unusual provisions which were entered into merely reflected the fact that the shares possessed a special value if sold to a dealer in shares. I cannot accept these submissions. It would be a complete delusion to regard the transaction in this case as a share dealing transaction coming within the area of trade of a dealer in shares. It was something very different. I would dismiss the appeal.”

[344]Lord Guest agreed with Lord Morris (at 623 B to H). He said that there was no inconsistency between Harrison and Finsbury. The decision of the majority in Harrison was that the mere fact that a transaction was entered into with the purpose of making a loss with no hope of making a profit and with a fiscal motive did not prevent it from being a trading transaction. The only transaction in question was the purchase of shares by a dealer who intended to do a dividend-strip. In the Finsbury case, the House considered that “by reason of the whole nature of the transaction it could not be a trading transaction”. He concluded, having read the speech of Lord Morris where he analysed the features of the transactions, that:

“I think they show clearly that the transactions in which the appellants were engaged were not the transactions of a normal trade in share dealing. The shares were not bought as stock in trade of a dealer in shares but as pieces of machinery with which a dividend-stripping operation might be carried out.”

[345]Lord Simon, Lord Donovan and Viscount Dilhorne considered that Finsbury and Harrison could not be reconciled and made their decision on the basis that the purpose or objective of the transaction in Lupton was to obtain a tax benefit.

[346]Viscount Dilhorne said, at 625 I to 626 A, that this was not a case where the appellants sought to make a profit by the exercise of their trade. The sale agreement shows that that was not contemplated. In his view:

“It would be wrong to hold that the shares were acquired for the purpose of and in the course of carrying on their trade unless it was established that the implementation of schemes for extracting money from the Revenue forms part of the trade of a dealer in stocks and share.”

[347]In considering Harrison and Finsbury he agreed, at 627 D to E, with Lord Morris’ statement that a trading transaction does not cease to be such merely because it is entered into in the confident hope that some fiscal advantage will result. If a transaction is established to be a trading transaction, it does not lose its character in consequence of a fiscal advantage. However, he thought that the question in Harrison was not whether a trading transaction ceased to be one but whether it was a trading transaction in the first place. He said, at 628 D, that he did not understand the rationale for Lord Morris describing Finsbury as “a wholly artificial device remote from trade to secure a tax advantage” when it was “no different” to Harrison. He said that he could not see “that the two cases are really distinguishable”.

[348]He noted, at 628 F to G, that in Finsbury Lord Morris attached importance to the position of the sellers of the shares as, unlike in Harrison, they were to receive a share of the tax recovered. However, he could not see how that could alter or affect the nature of the purchaser’s activity. A dealer trading in stocks and shares “does not cease to trade in them if he agrees to share with the vendor as part of the purchase price any profit he makes”. If there is no valid ground for distinguishing between the two cases, “the choice must lie between following Harrison or Finsbury,” in which case he would “unhesitatingly follow Finsbury.”

[349]Applying that approach, he concluded, at 628 H to 629 A, that:

“… if a transaction viewed as a whole is one entered into and carried out for the purpose of establishing a claim against the Revenue … I for my part would have no hesitation in holding that it does not form part of the trading activities of a dealer in stocks and shares. When I say ‘viewed as a whole’ I mean that regard must be had not only to the inception of the transaction, to the arrangements made initially, but also to the manner of its implementation.”

[350]Lord Donovan noted, at 629 B to D, that “the ordinary trader in stocks and shares normally makes his purchases on the attractions of the investments as a merchantable commodity” but the purchase of shares in this case was not decided upon by the purchaser as the result of any such commercial appraisement. They were bought:

“pursuant to a plan having as its objects (a) to provide the Gill family with the equivalent in capital of certain undistributed profits which if taken by way of dividend would attract surtax; and (b) to provide the appellants with an opportunity to compel the Revenue to pay to them a large sum of money which they, the appellants, had never themselves disbursed in tax, and which on recovery they would share with the vendors of the shares.”

[351]He concluded, at 629 E to G, “that this is not trading in stocks and shares”. If he was asked what it is, he would reply that it is:

“the planning and execution of a raid on the Treasury using the technicalities of revenue law and company law as the necessary weapons. There is an obvious distinction between this kind of case and the case where fiscal advantages are incidental.”

[352]He continued that the House of Lords drew that distinction in Finsbury when confronted with Harrison. That was also a case where “the fiscal advantage was the sole objective”; and it is not easy, therefore, to see why it should be treated as the opposite of Finsbury. He said that in Harrison, as he read it, the majority view was reached by examining the component parts of the transaction, (the purchase of shares, the receipt of a dividend and the sale of the shares) and proceeding thence to the conclusion that these things, when done by a dealer in shares, amount to trading in shares. It was held to be irrelevant that such trading had a fiscal advantage in view. The minority view did not “confine itself to examining the ingredients of the transaction” but recognised that “its totality may be different from the mere sum of its parts” so that when the transaction was viewed as a whole “it stood revealed as a device for extracting money from the Exchequer and nothing else”.

[353]He said, at 629 H, that in the Finsbury case the component parts of the transactions if considered alone would logically have produced the same decision as in Harrison. But there the House had taken “a comprehensive view of the transaction as a whole; and decided that it was “a wholly artificial device remote from trade to secure a tax advantage”. In his view, it is immaterial in principle that the wider view was induced by certain unusual features in Finsbury. The “altered approach”, with which he agreed, “must now clearly be taken to be right.”

[354]Lord Simon said, at 631 A to B, that, adopting the approach of reading Harrison in the light of Finsbury, “it is clear that the former was a very narrow decision indeed, and that “particular caution is required in the use of passages from the speeches in the former case which cannot be reconciled with the decision (or, indeed, the language of judgment) of the latter”. In his view, the two cases, taken as a binary system, establish the following propositions, at 631 B to E: “(1) the question in every case is whether the relevant loss has been incurred in the course of trade (of dealing in shares); (2) dividend-stripping (or any other transaction to secure a fiscal advantage) is not in itself part of the trade of dealing in shares (see also Inland Revenue Commissioners v. Dowdall, O' Mahoney & Co. Ltd. [1952] A.C. 401); (3) share-dealings and other business transactions vary almost infinitely; and to determine whether the transaction is, on the one hand, a share-dealing which is part of the trade of dealing in shares or, on the other, merely a device to secure a fiscal advantage, all the circumstances of the particular case must be considered; (4) a share-dealing which is palpably part of the trade of dealing in shares will not cease to be so merely because there is inherent in it an intention to obtain a fiscal advantage, or even if that intention conditions the form which such share-dealing takes; (5) what is in reality merely a device to secure a fiscal advantage will not become part of the trade of dealing in shares just because it is given the trappings normally associated with a share-dealing within the trade of dealing in shares; (6) if the appearance of the transaction leaves the matter in doubt, an examination of its paramount object will always be relevant and will generally be decisive (see also Iswera v. Inland Revenue Commissioner

[1965] 1 WLR 663

, especially at p. 668).

[355]He continued, at 631 F to G, that he agreed with the analysis and description of the transactions set out by Viscount Dilhorne and: “Such trappings of the trade of dealing in shares as we have here are quite inadequate to prevent the real nature of this transaction showing through”.

[356]He referred to the provisions in the sale agreement which related to the vendor sharing in the tax reclaim arrangements and concluded, at 632 B to C, as follows:

“My Lords, this was not share-dealing within the trade of dealing in shares. It is plainly a joint venture of the appellants and the vendors of the shares, by taking advantage of quirks of revenue and company law, to obtain money out of the public purse and share it between them. Even if the transaction were equivocal, its true nature would, in my view, be resolved by investigation of its paramount object: since, on the findings of the special commissioners, the transaction would produce a loss to the appellants unless repayment of income tax were obtained, I conclude that the paramount object of the transaction was to procure such repayment of income tax: it was, in other words, a tax-recovery device.”

[357]In Gurneville, the facts were somewhat more complicated than in Lupton but the way the scheme operated was essentially the same. The scheme was structured, however, so that the share dealer made a commercial profit regardless of the income tax repayment. The taxpayer placed a great deal of emphasis on this feature but none of their Lordships thought that the intention to make and the actual making of this profit sufficed for the transaction to be regarded as a dealing in shares.

[358]In summary, a company, BI, owned a number of subsidiaries which held over 100 properties standing at a gain. The subsidiaries transferred the properties at book value to another group company, BP, which was a land dealer. BP sold off many of the properties and, as it did so, paid the resulting profits to BI by way of dividends. An unconnected share dealer, agreed to buy the shares in BI for an initial price of £16,685 but to be adjusted by reference to the net assets at a specified date following completion initially at a discount of 5% which was later adjusted to 6%. The net asset adjustment date was set to coincide with the date BP was expected to cease trading once it had sold all the properties. Following the purchase of the shares in BI, the profits then in BI were paid to the dealer by way of dividend. The dealer wrote down the value of the shares in BI, claimed it made a loss and claimed a repayment of tax. Due to it paying a discounted price under the net asset mechanism, the dealer realised a profit of £90,996 on the eventual sale of the shares.

[359]Lord Morris said, at 670 G, that “each case must depend upon its own facts and decision can only be reached when all the facts are surveyed”. Only then can “the shape and structure and nature of what has been created be seen in perspective”. He noted, at 671 F, that the scheme was “designed to produce financial profit which irrespective of fiscal advantage could be described as a commercial profit.” Given how the price was to be calculated, at 672 A, it “is fair to suppose that this gave the purchasers a very confident assurance of profit even if their expectations of fiscal advantages did not mature”. He continued, at 673 B to C, that the arrangements which were made “resulted in financial benefit both for the sellers of the shares and for [the dealer] and that such benefit accrued because fiscal matters were handled with great acumen”. Using similar terms to those he used in Lupton he said that the transactions:

“were not merely inspired by fiscal considerations: the provisions in regard to fiscal matters which for mutual benefit were calculated to produce financial advantage were part of the pith and substance of the transactions themselves.”

[360]At 673D, Lord Morris concluded that, in determining the question of whether there was a trading activity of a dealer in shares, it may be of no account “whether a transaction is a usual or an unusual one or whether it is a simple one or is complicated”. Rather, again taking a similar approach as in Lupton, the question is:

“… whether the transaction bears the stamp and mark of the trade of a dealer in shares or whether its very structure and content reveals it as something different in kind. Approaching the enquiry on the lines that I explained in my speech in Lupton ... I have no doubt that the transactions now under review were not those that can be regarded as trading transactions in the course of their trade of dealers in shares.”

[361]Viscount Dilhorne noted, at 674 I, that the taxpayer strenuously contended that the fact that a profit was made showed that the transaction had a commercial character and stressed that the features present in it were those ordinarily to be found when such a dealer is trading. He agreed that “the purchase of shares, their sale and the receipt of dividends are common features of such a dealer’s trade” but “so they are of a dividend stripping operation and one designed and planned to secure sums from the Revenue” and, therefore, as he had said in Lupton, one must carry out a wider exercise, at 675 A:

“One must look further to determine the true character of the transaction and, looking at it as a whole, one is entitled to have regard to the fact that the profit of £90,996 was far less than the amount initially claimed from the Revenue, £413,706.”

[362]He concluded, at 675 D to E, that:

“Looking at the transaction as a whole, the conclusion is I think inescapable that it was one designed, intended and carried out, so far as the Respondent Company was concerned, mainly to provide a basis for claims against the Revenue. Whether all the assets of BI could have been distributed by way of dividend I do not know, but the fact that £90,996 was not does not in my opinion alter the character of the operation. Such an operation was not in my opinion one which came within the scope of the Respondent Company’s trade.”

[363]Lord Donovan reached similar conclusions. At 675 G to H, he said that it was “plain that the transaction was part of a scheme” whereby the sellers of the shares would receive capital profits which if received as dividends would attract surtax and the taxpayer would be able to enrich itself by the device of dividend stripping. He concluded that:

“when shares are bought for the sole or main purpose of dividend stripping, the transaction is not a trading transaction.”

[364]He repeated what he had said in Lupton and, in particular, noted that he was still not able to “to perceive any line differentiating in essentials the case of Harrison ... from the case of Finsbury.” He said, at 676 A to E, that his conclusion was not affected by the fact that the taxpayer intended from the start to realise a profit and did make that profit. The weight which would otherwise attach to this circumstance was diminished by a number of factors.

[365]The provision for such a profit was “part and parcel of the scheme”, as the taxpayer was to pay a price for the shares of BI which was 5 to 6% less than the value of its underlying assets. It was “difficult to separate out this element of the scheme and assess its weight as an independent feature”. Any such assessment must necessarily be made as at the date of the inception of the scheme; and, at that time, it must have been distinctly speculative. By contrast “the highly skilled persons advising all those engaged in the scheme would have been able to make a fairly shrewd forecast of the upper and lower limits of the profits to be expected…”. He noted that the taxpayer pointed to the magnitude of the commercial profit but that the tax claims arising out of the dividend stripping came to some £430,000. He concluded that:

“Looking at the matter as a whole I do not think the scheme takes on a different colour because the intention of those behind the Respondent Company was to make a commercial profit as well. Predominantly its aim was to exploit certain features of the existing fiscal system.”

[366]Lord Simon said, at 677 I, that although the fiscal advantages to both the sellers and the taxpayer are manifest, that did not necessarily mean that this was not a trading transaction as it was necessary to look at all the circumstances:

“It is only by examining all the circumstances of the case that it can be determined whether the transaction was, on the one hand, a share dealing in the course of the trade of dealing in shares or, on the other, a mere device to secure a fiscal advantage.”

[367]He said, at 678 G, that the profit was a “commercial profit of a very curious nature”:

“It was built into the scheme from the outset by the stipulation that the price to be paid should be fixed at 95 per cent. (later 94 per cent.) of the value of the underlying assets as subsequently ascertained. The result was that on resale the Respondents were bound to make a profit…”

[368]He thought it likely, at 678 H to I, that the profit should be taken to be a “reward for [the taxpayer] putting up to the vendors the scheme of tax avoidance”. If that related solely to the advantages accruing to the vendors, the 5 to 6% reward was in no sense a commercial profit earned by the respondents in their trade of dealing in shares. On the other hand, if as he thought more likely on the evidence, the whole scheme with tax advantages both for the sellers and the taxpayers was put up to the vendors, then it could be regarded as, at 679 A to B:

“either a reward to the respondents for putting up a tax avoidance scheme ... or, being an integral part of the entire scheme, as a colourable device to make a mere expedient for extracting money from the public purse appear to be a bona fide dealing in shares, by assuming the semblance of a ‘commercial profit’ arising from the transaction; or partly the one, partly the other. On no view was it true ‘commercial profit’. It would be absurd, moreover, to remain oblivious to the quantitative relationship of the respective advantages, ‘commercial’ or ‘fiscal’. The ‘commercial profit’ was £90,996: the total claim against the Revenue amounted to £413,706….”

[369]In conclusion, at 679 D to E, he referred back to his speech in Lupton as to the question to be answered in this type of case:

“whether, in light of all the circumstances, the transaction is, on the one hand, a share dealing which is part of the trade of dealing in shares (albeit intended to secure a fiscal advantage, or even conditioned in its form by such an intention) or, on the other, a mere device to secure a fiscal advantage (albeit given the trappings normally associated with a share dealing within the trade of dealing in shares).”

[370]In this case, at 679E to F, he thought the question could be narrowed but that, however the question was formulated, the answer was that this was merely a device to secure a fiscal advantage:

“looking at the transaction as a whole, was it, on the one hand, one whereby a true commercial profit was taken in a fiscally advantageous way or, on the other, one in which a ‘commercial profit’ was merely a by-product of, or a disguise for, what was really a tax recovery device? Whichever way the question is put, I have no doubt that, judged both quantitatively and qualitatively, the transaction falls into the latter category in each case.”

Coates v Arndale and Nova Securities

[371]The cases of Coates (Inspector of Taxes) v Arndale Properties Ltd

[1984] WLR 1328

(‘Coates v Arndale’) and Reed (Inspector of Taxes) v Nova Securities

[1985] 1 WLR 193

(‘Nova Securities’) [1985] UKHL TC 59 516 both relate to provisions which enable a taxpayer who appropriates an asset to trading stock to convert a capital loss into a trading loss. The issue was whether the relevant asset was acquired as trading stock which required, as Lord Templeman said in Nova Securities, that the asset must be not only of a kind which is sold in the ordinary course of the company’s trade but must also be acquired for the purposes of that trade with a view to a resale at a profit. The cases raise a similar issue, therefore, to that in the current case.

[372]In Coates v Arndale, one of a number of group companies, SPI, had developed an investment property at a cost of £5.3 million but its market value was only £3.1 million. Arrangements were put in place to convert the potential capital loss standing in SPI into a trading loss which could be group relieved. SPI sold the property to another subsidiary, Arndale, which was a dealer in property. Arndale then sold it to another subsidiary, APTL, which was to hold the property on capital account for its market value of £3.1 million. Arndale made a £10,000 profit on the transaction as SPI sold the property to Arndale at £10,000 below its open market value of £3.1 million.

[373]Arndale claimed it had acquired the property as trading stock. On that basis it was permitted under the legislation relating to the appropriation of trading stock to compute its trading profits on the basis that it had incurred the same level of expenditure as SPI, thereby realising a trading loss, which could be surrendered by way of group relief. The House of Lords, however, rejected this. Lord Templeman said at 557:

“In my opinion Arndale never decided to acquire, and never did acquire, the lease as trading stock. The group’s advisers procured the transfer of the lease from SPI to Arndale and from Arndale to APTL with the object of obtaining group relief of £2.2 million trading loss without in fact changing the lease from a capital asset to a trading asset. The group seeks the advantage of treating the lease as trading stock while ensuring that the group retains the lease as a capital asset at all times. Arndale followed instructions and lent to the transaction its name and its description as a property-dealing company. Arndale did not trade and never had any intention of trading with the lease.”

[374]He went on to state, at 557, that, in effect, the £10,000 was not truly a commercial profit for entering into two assignments of property worth over £3 million but rather it was provided “in order to give the whole transaction a faint air of commercial verisimilitude”:

“The award of £10,000 was ostensibly made at the expense of APTL which paid Arndale for the lease £10,000 more than the price paid by Arndale to SPI. In reality the award of £10,000 was made at the expense of SPI which sold for £10,000 less than the market value assessed by the group. The profit of £10,000 did not represent the difference between the price at which Arndale negotiated the purchase and the price at which Arndale negotiated the sale. The profit of £10,000 did not represent the difference between the value of the lease to SPI and the value of the lease to APTL. The profit of £10,000 was a timid veil designed to conceal the fact that the lease was not being traded. Moreover, all three companies being wholly owned subsidiaries of the same parent, the £10,000 was a book entry which had no material effect on the overall financial position of the group.”

[375]In Nova Securities, the Littlewoods group acquired shares and debts which were standing at a substantial capital loss. With a view to converting the loss into a trading loss which could be group relieved, the group acquired a share dealing company to which the debts and shares were transferred. As in Coates v Arndale, the trading company claimed to make a trading loss under the provisions relating to the appropriation of assets to trading stock. Again, the leading judgment was given by Lord Templeman. He agreed, at 563 F, with the Court of Appeal that for an asset to be acquired as trading stock “it must be not only of a kind which is sold in the ordinary course of the company’s trade but must also be acquired for the purposes of that trade with a view to a resale at a profit”.

[376]He noted, at 564, that, in the Court of Appeal, both Lawton LJ and Fox LJ relied on the principle to be deduced from the dividend-stripping cases and particularly on the decision in Lupton. He described Lupton as deciding that “relief was not to be granted because the object of the transaction was to obtain a fiscal advantage and the purchase did not form part of the trading activities of a dealer in stocks and shares”. He referred to the comments of Viscount Dilhorne, as set out at [349] above, and of Lord Simon of Glaisdale, as set out at [354] above that, “what is in reality merely a device to secure a fiscal advantage will not become part of the trade of dealing in shares just because it is given the trappings normally associated with a share-dealing within the trade of dealing in shares;..” He concluded that in a dividend-stripping case, such as Lupton, “an artificial loss is artificially created and the artificial transaction does not constitute trading but constitutes the manufacture of a tax advantage”. In the present case, however, the Littlewoods group sustained a real loss.

[377]He went on to note, at 565, that whilst the legislature has not extended group relief to allowable losses, it has conferred power on a group of companies to convert an allowable loss into a trading loss which can then be “shuffled to secure a tax advantage”. He did not see that there could be a cause for complaint that a company seeks to benefit from a fiscal advantage provided for by the legislation. The only requirement was that there must be an acquisition by a trading company as a trading stock.

[378]So far as the bank debts were concerned, he agreed, at 565, with Fox LJ in the Court of Appeal (

[1984] 1 WLR 537

, 554) that the property was acquired by Nova “as trading stock” (essentially on the basis that there were commercial justifications for the acquisition). However, the shares were worthless such that there was simply no commercial justification for their acquisition and there was “no conceivable reason, apart from section 274, why the shares should change hands at all…But section 274 only applies to the shares if the shares were acquired by Nova as trading stock, namely with a view to the resale of the shares at a profit. The shares were not commercially saleable at any price.” Ensign Tankers

[379]In Ensign Tankers, the taxpayer company and four other UK companies entered into a limited partnership (the Victory Partnership) as limited partners with the subsidiary of an American film production company as general partner to produce and exploit the film “Escape to Victory”. The object of the UK companies was to claim first year tax allowances in respect of the whole of the cost of the film of $14 million. However, under the agreement with the production company, the partnership contributed $3.25 million (of which $2.3 million represented the taxpayers’ contribution) towards the cost of the film. Under the agreement, the partnership became entitled to the master negative of the film, the production company agreed to complete the film and to lend to the partnership the cost of making the film in excess of the partnership contribution. The loans were non-recourse and were only repayable out of 75% of the net receipts from the exploitation of the film. The remaining 25% of the net receipts was received by the partnership. The taxpayer appealed against the refusal of the inspector of taxes to allow its claim to first year allowances on its expenditure on the film which was made by reference to the full cost of $14 million.

[380]The issue, as identified by Lord Templeman at the start of his judgment, was that the taxpayer claimed for itself and its partners capital allowances for expenditure of $14 million although they were never liable to spend more than $3.25 million of their own money. He noted at the outset, at 661 C, that this was “a tax avoidance scheme, a single composite transaction whereunder the tax advantage claimed by the taxpayer is inconsistent with the true effect in law of the transaction”. Lord Templeman considered the financial consequences to be clear. Victory Partnership incurred expenditure of $3.25 million. It had a right to 25% of the net receipts from the exploitation of the film whereby it received $3 million. The taxation consequences were that Victory Partnership, provided it were trading, generated a first year allowance of $3.25 million and Victory Partnership became in due course liable to corporation tax on the profits of $3 million which it received. But the non-recourse nature of the borrowing ensured that the production company paid the whole cost of the film exceeding $3.25 million and, conversely, that Victory Partnership would not be liable for the cost of the film in excess of $3.25 million.

[381]Lord Templeman noted the rival arguments at 668 H to 669 A.(1) The appellant submitted that: “the taxpayer may enter into any transaction in any form he pleases and the court is confined to that form and cannot have regard to the rights and obligations which flow from the transaction because the court cannot consider the substance of the transaction”.

(2) The Revenue on the other hand appeared: “to look upon tax avoidance as a corporate cancer which infects and `destroys any fiscal effect advantageous to the taxpayer. In the present case although the contribution by Victory Partnership to the cost of the film of the sum of $[3.25m] in consideration for 25 per cent. of the net receipts from the exploitation of the film can only be described as trading, Victory Partnership did not generate a first year allowance of $[3.25m] because the trading was part of a tax avoidance scheme designed to procure a first year allowance of $14m”.

[382]He continued to set out the cases on tax avoidance schemes including an outline of the decision in Lupton which the Revenue particularly relied on. He described, at 670 C to D, the approach taken in that case as being that: “the House considered the scheme as a whole and declined to allow the taxpayer a fiscal result which did not correspond to the financial result…..In Lupton’s case the dealer was not allowed to succeed in a claim for a fiscal loss …. because, viewing the transaction as a whole, and taking the dividend into account he had made no loss at all”.

[383]He referred, at 670 E to G, to the same passage from Viscount Dilhorne’s judgment as he had referred to in Nova Securities (see [376] above) and to Lord Donovan’s comments, as set out at [350] to [351] above, that that case involved the “planning and execution of a raid on the Treasury”. He noted, at 670 H, that the revenue, relied heavily on Lupton because it was held in that case that “tax avoidance is not trading”. Therefore, they submitted, the tax avoidance scheme was not trading and the partnership did not create any valid claim to a first year allowance although the partnership incurred expenditure of $3.25 million in the production of a film.

[384]Lord Templeman continued, at 671 A, to say that he could see the force of the argument and noted that Lupton was followed by the Court of Appeal in the instant case. However, at 671 A to B, he did not think Lupton determined this case because in that situation the trader simply did not trade at all:

“But in dividend-stripping cases the tax avoidance scheme negatives trading because on the true analysis of the transaction the trader does not trade at all. In Lupton’s case where there was neither a profit nor a loss the House did not consider the present situation in which on the true analysis there was trading involving an expenditure of $[3.25m]. The financial consequences of the scheme, namely the expenditure by Victory Partnership of $[3.25m] on the making of a film, produce the corresponding fiscal consequence of a first year allowance of that sum. The task of the courts is to construe documents and analyse facts and to ensure the taxpayer does not pay too little tax or too much tax but the amount of tax which is consistent with the true effect in law of the taxpayer’s activities. Neither the taxpayer nor the revenue should be deprived of the fiscal consequences of the taxpayer’s activities properly analysed.”

[385]He thought that the Special Commissioners had taken the wrong approach by simply disregarding all the fiscal consequences of the transaction on the basis that it was a tax avoidance scheme with fiscal motives as the paramount object. In his view, at 676 H to 677 A, the Ramsay line of cases did not authorise that approach:

“The principles of Ramsay and subsequent cases do not authorise the court to disregard all the fiscal consequences of a single composite transaction read as a whole on the grounds that it appears that the transaction is a tax avoidance scheme. In the present case the commissioners felt bound to ignore all the fiscal consequences which are beneficial to the taxpayer because Victory Partnership had entered into the scheme with ‘fiscal motives as the paramount object’. ”

[386]He said, at 677 A, that similarly Sir Nicolas Browne-Wilkinson V-C had applied the wrong approach in the Court of Appeal in holding that “the taxpayer is deprived of all the beneficial effects of the scheme if the scheme was entered into essentially for the purpose of obtaining a fiscal advantage under the guise of a commercial transaction:

[1991] 1 WLR 341

, 357.”

[387]He then quoted, at 677 C, the following from the Vice-Chancellor’s decision (taken from page 355 of that decision):

“if the commissioners find as a fact that the sole object of the transaction was fiscal advantage, that finding can in law only lead to one conclusion, viz. that it was not a trading transaction. ... if the commissioners find as a fact only that the paramount intention was fiscal advantage . . . the commissioners have to weigh the paramount fiscal intention against the non-fiscal elements and decide as a question of fact whether in essence the transaction constitutes trading for commercial purposes.”

[388]He then criticised that approach in the following passage, at 677 D to F:

“My Lords, I do not consider that the commissioners or the courts are competent or obliged to decide whether there was a sole object or paramount intention nor to weigh fiscal intentions against non-fiscal elements. The task of the commissioners is to find the facts and to apply the law, subject to correction by the courts if they misapply the law. The facts are undisputed and the law is clear. Victory Partnership expended capital of $[3.25m] for the purpose of producing and exploiting a commercial film. The production and exploitation of a film is a trading activity. The expenditure of capital for the purpose of producing and exploiting a commercial film is a trading purpose. By section 41 of the Act of 1971 capital expenditure for a trading purpose generates a first year allowance. The section is not concerned with the purpose of the transaction but with the purpose of the expenditure. It is true that Victory Partnership only engaged in the film trade for the fiscal purpose of obtaining a first year allowance but that does not alter the purpose of the expenditure. The principles of Ramsay and subsequent authorities do not apply to the expenditure of $[3.25m] because that was real and not magical expenditure by Victory Partnership. The Vice-Chancellor referred to authorities in which intentions sometimes illuminated and sometimes obscured the identification of a trading purpose. But in every case actions speak louder than words and the law must be applied to the facts.”

[389]Lord Templeman then set out, at 677 H to 679 H, a number of cases which it appears were intended to illustrate that intentions sometimes illuminated and sometimes obscured the identification of a trading purpose including Coates v Arndale and Nova Securities. Following his review of the cases he concluded, at 680 A to C, that:

“In the present case the legal effect of the transaction, whatever its design was a trading transaction whereby Victory Partnership expended $[3.25m] towards the production of a film in which Victory Partnership had a 25 per cent interest. All these authorities were dealing with the identification of a trading transaction. In the present case a trading transaction can plainly be identified. Victory Partnership expended capital in the making and exploitation of a film. That was a trading transaction which was not a sham and could have resulted in either a profit or a loss. The expenditure of $[3.25m] was a real expenditure. The receipts of $3m were real receipts. The expenditure was for the purpose of making and exploiting a film and entitled Victory Partnership to a first year allowance equal to the expenditure. The receipts imposed on Victory Partnership a corporation tax liability.”

[390]Finally, we note that Lord Goff and Lord Jauncey adopted different reasoning but reached the same conclusion. Lord Goff also concluded, at 684 C, that the partnership was trading, though only to the extent of its investment of $3.25 million and no more; that sum constituted, on a true construction of the statute, the only capital expenditure incurred by the partnership in the making of the film. He said that this conclusion lead to the “sensible and realistic consequence that the partnership was not deprived of a capital allowance in respect of that sum, which would have been the case if the partnership had been held not to have been trading at all.”

[391]Lord Jauncey considered, at 684 F onwards, whether the transaction should rather be treated as a non-trading one in its entirety looking at the decision in Lupton. He concluded, at 685 D to H, that was not the case. He said that Lupton was “an all or nothing case in which the only question was whether or not the relevant transactions formed part of the trading activities of a dealer in stocks and shares”. Lupton was not concerned with what would otherwise be a trading transaction producing financial and fiscal consequences that formed part of a tax avoidance scheme: “I do not consider that [Lupton] requires that the trading transaction be denatured because the taxpayer has incorporated it within a tax avoidance scheme which seeks to obtain for him greater fiscal advantages than the trading transaction if standing alone would produce. When Parliament has provided that a taxpayer shall be entitled to certain allowances in certain circumstances I can see no reason in principle why when those circumstances exist he should be deprived of those allowances simply because he has sought and failed to engineer a situation in which he obtained allowances greater than those to which the circumstances entitled him. Where, as here, there is, as my noble and learned friend has pointed out, an end result which has both financial and fiscal consequences, the proper approach is to disregard the steps in this scheme which have no commercial purpose rather than to treat those steps as somehow affecting or denaturing other steps in the scheme having such a purpose. In support of this proposition I cannot do better than cite the well known passage from the speech of Lord Brightman in Furniss v. Dawson

[1984] AC 474

, 527”. New Angel Court

[392]New Angel Court is another case where the issue was the applicability of the trading stock appropriation rules. The group had a number of properties which were standing at a loss of some £68 million for capital gains purposes. With a view to converting the potential capital loss into a trading loss (which could be group relieved) the relevant company transferred the properties to another group company, NAC, which carried on the business of developing and dealing in property. NAC claimed it acquired the properties as “trading stock” (within the meaning of s 173(1) of the Taxation of Chargeable Gains Act 1992) so that they were appropriated at their historic cost meaning that a loss was generated. NAC sold the properties to third parties for a commercial profit. The lower courts decided that the properties were not acquired as trading stock as there was no purpose for the transfer other than achieving the tax benefit.

[393]Having set out a comprehensive review of the cases, Jonathan Parker LJ noted, at [87], that by enacting s 173(1), Parliament has provided groups of companies with an opportunity to secure an advantage for corporation tax purposes by appropriating an asset as trading stock and thereby converting a capital loss of one company in the group into a trading loss available for group relief. Accordingly, “as Lord Templeman observed in Reed v. Nova Securities …, the Revenue cannot complain that a taxpayer has obtained a tax advantage by availing itself of the opportunity which the legislation itself offers”. He said that it was in that context and against that background that the significance of the Special Commissioners’ findings as to the purpose behind the transfer of the properties to NAC had to be considered and, in particular, their finding that “there was no purpose other than tax for the transfer”.

[394]However, before turning to the Special Commissioners’ findings he made some general points. The first was, at [89], that, in the light of Lord Templeman’s speech in Ensign Tankers, the observations of Sir Nicolas Browne-Wilkinson V-C in that case relating to the relevance of fiscal purpose, and “in particular the distinction which he sought to draw between sole purpose and paramount purpose, are no longer good law”.

[395]The second, at [90] and [91], was that he thought it clear from Coates v Arndale and Reed v Nova Securities, and confirmed by Lord Templeman in Ensign Tankers that, in determining whether an asset has been acquired “as trading stock” for the purposes of section 173(1), “fiscal considerations are to be ignored.” He noted that in both of those cases Lord Templeman was content to assume that the group’s object was to obtain a tax advantage and, at [92], that in Ensign Tankers Lord Templeman was able to conclude that “the composite transaction in that case was (at least in part) a trading transaction, ‘whatever its design’.”

[396]In his judgment, at [93] and [94], fiscal considerations were entirely to be put to one side in deciding whether an asset is acquired as trading stock for the purposes of the appropriation rules; in such cases, a taxpayer is merely seeking to obtain a fiscal benefit provided for by the legislation itself:

“the mere fact that a group of companies sets out to avail itself of the opportunity of obtaining a fiscal advantage which Parliament has itself provided says nothing as to whether the requirement which Parliament has imposed as the condition of obtaining that fiscal advantage – that is to say that the asset in question must be acquired ‘as trading stock’ – has been fulfilled……... So in my judgment fiscal considerations (whether they be described in terms of motive, purpose, or object) must be put entirely on one side in considering whether an asset was acquired ‘as trading stock’ for the purposes of section 173(1).”

[397]He went on to say, at [95], that it is plain from Coates v Arndale and Reed v Nova Securities that an investigation is required into purpose in the sense that it is necessary to consider whether the asset was acquired by the taxpayer “for the purposes of that sale with a view to resale at a profit”. He noted that in Coates v Arndale that was why there was no allowable loss as regards the acquisition of the lease by Arndale. He concluded, at [98], and [99], that:

“Section 173(1) does not require the absence of fiscal considerations as elements in the acquisition of the asset in question: rather, it requires the presence of a trading purpose. As Coates v Arndale and Reed v Nova Securities demonstrate, a trading purpose is not negatived by the presence of fiscal considerations: to use Megarry J’s word (in Lupton) the existence of fiscal considerations will not ‘denature’ a trading purpose, just as the existence of fiscal considerations will not prevent what would otherwise be a trading transaction from being regarded as such for the purposes of section 173(1). At the heart of the matter, as it seems to me, is the need to recognise that in the context and for the purposes of section 173(1) a trading transaction may be dictated entirely by fiscal considerations, without losing its character as a trading transaction.”

[398]He considered, at [102], that the Special Commissioners erred in principle in so far as they found that the group’s purpose of disposing of the properties, whichever company owned them, somehow negatived the existence of the trading purpose which (implicitly) they would have found to exist “looking at [NAC] in isolation”. It followed from his earlier conclusion that the fact that the group decided to avail itself of the benefit provided for by s 173(1) and that the group had adopted a policy of selling off its commercial properties says nothing as to whether NAC acquired the properties as trading stock.

[399]He concluded, at [110], that NAC acquired the properties as “trading stock”:

“As Lord Templeman said in Ensign Tankers……. ‘actions speak louder than words, and in every case the law must be applied to the facts’. The facts in the instant case were that (as the Special Commissioners found) the price at which NAC acquired the Properties was a proper market price; that following its acquisition of the Properties NAC set about selling them and succeeded in selling all but one of them; and that in so doing NAC made a real profit (contrast the ‘timid veil’ in Coates v Arndale). The conclusion follows that NAC’s acquisition of the Properties was a trading transaction entered into for a trading purpose; and that NAC accordingly acquired the Properties ‘as trading stock’ within the meaning of section 173(1).”

Discussion on case law

[400]BNP argued that Lupton established that a given transaction must, in the first instance, be assessed objectively in the light of what the taxpayer actually did without regard to the fiscal consequences of the transaction or to motive (whether fiscal or otherwise). On that analysis:(1) If, on an objective assessment, the transaction is plainly part of an established trade of dealing in the type of asset in question, it is a trading transaction.

(2) If, on an objective assessment, the transaction is plainly not part of an established trade of dealing in the type of asset in question, it will not qualify as a trading transaction. That would be the case if the fiscal element has invaded the transaction itself, moulding and explaining it; even if it is given the trappings normally associated with trading transactions of the type in question.

(3) It is only if the objective features of the transaction are ambiguous or equivocal, that reference to motive can be made.

[401]BNP asserted that all their Lordships in effect decided that Lupton was in the second category where, on an objective analysis, the taxpayer was unambiguously not carrying out a trading transaction. The focus was very much on the particular transaction carried out by the taxpayer as discerned from the documentary terms on which it was entered into. The terms were such that it was just not capable of being described as a share dealing transaction, rather, they revealed it to be a joint venture between vendor and purchaser to recover money from HMRC. Their Lordships were clearly influenced by the “singular” and “remarkable” features in the sales agreement which meant that, under the terms of the transaction itself, the sellers potentially shared in the tax repayment obtained by the dealer. It is easy to see why that was not regarded as a trading transaction on the basis that the fiscal element had invaded the transaction itself, moulding and explaining it (using the terms used by Megarry J in the High Court as quoted with approval by Lord Morris). Applying an objective analysis to the transaction in question here, on the contrary, according to its terms, the transaction is plainly carried out in the course of the London branch’s trade.

[402]The subsequent decision of the House of Lords in Gurneville is just an application of the principles established in Lupton; it does not extend those principles. Again, there were a number of unusual features which were of relevance (as dictating the design of the transaction) which are not present in this case. As in Lupton, the “provisions in regard to fiscal benefit” were in the agreements themselves.

[403]HMRC pointed to the judgments of Viscount Dilhorne, Lord Simon of Glaisdale and Lord Donovan in Lupton as representing the majority view. They noted that they each preferred the broader Finsbury view of looking at the transaction “viewed as a whole” in forming their conclusions that the transaction was merely a device for securing a tax repayment or a raid on the revenue with merely the trappings associated with share dealing. HMRC considered that this is reinforced in Gurneville. They noted, in particular, that in taking the same approach, of looking at the whole circumstances, it was held in that case that the commercial profit did not affect the analysis; it was simply part of the design of the scheme (and was dwarfed by the tax profit). It was also, in their view, clear from that case that the assessment is not confined to the terms of the contract which effects the transaction. As it happened, on the facts of Lupton, it was possible to get the whole picture from the terms of the contract. But it does not follow that in more complicated cases, regard cannot be had to the broader arrangements. In their view, in this case, looking at the whole of the arrangements, as in Gurneville, the “commercial” aspects of this transaction were wholly engineered in order that a commercial profit could be generated which was tiny compared with the real benefit, the s 730 benefit, that the transaction was designed to achieve.

[404]Essentially, we agree with HMRC that the reasoning of Viscount Dilhorne, Lord Simon and Lord Donovan represents the majority view in Lupton and Gurneville. The difference in approach stemmed from the fact that the majority clearly saw the earlier dividend stripping cases of Harrison and Finsbury as conflicting decisions whereas Lord Morris and Lord Guest sought to reconcile them.

[405]Viscount Dilhorne and Lord Donovan considered the two cases were wholly irreconcilable because, as Lord Donovan noted, in each case the “sole objective was tax avoidance”. In their view the broader approach in Finsbury was to be preferred and the transaction should be viewed “as a whole” or “in its totality” which revealed dividend stripping as a fiscal device rather than the narrow component by component approach taken in Harrison. Lord Simon stopped short of quite such an overt criticism but cautioned that Harrison was narrow and reconcilable only in establishing a binary set of principles as he set out.

[406]In rejecting the component by component approach in Harrison, the majority took a broader realistic view of the facts which is not dissimilar to the modern approach to tax avoidance cases. In their view, the dividend stripping transactions were not part of a trade of share dealing because, on an objective assessment, looking at the whole of the circumstances, the arrangement was a device to obtain a tax benefit (as opposed to buying and selling shares for a profit). The transaction was designed and implemented with that object. As Viscount Dilhorne said, looking both at the “manner of inception and implementation” of the transaction, it was clear the transaction in Lupton was entered into “for the purpose of establishing a claim against the revenue”. Lord Donovan held that the transaction was a “plan” having “as its objects” the opportunity to recover tax which may be described as “the planning and execution of a raid on the Treasury”. It was not a case where “fiscal advantages are incidental”.

[407]Lord Simon similarly said that the question was whether the transaction was “merely a device to secure a fiscal advantage” which had to be considered in “all the circumstances”. A share dealing which is “palpably part of the trade” in his view would not cease to be such because there “is inherent in it an intention to obtain a fiscal advantage” but what is “in reality merely such a fiscal “device” would not be part of a trade “just because it is given the trappings normally associated with a share-dealing”. If the appearance of the transaction leaves the matter in doubt, an examination of its “paramount object” had to be made. He thought the transaction was plainly a tax recovery device but even if it were equivocal its “paramount object” was as such (as it would be loss making but for the tax recovery).

[408]Lord Morris (and Lord Guest who agreed with him) continued to justify the different conclusion reached in Harrison and Finsbury essentially on the basis that in Harrison there was no tax element which intruded into the terms of the bargain made between the parties. So a transaction which has “all the essential hallmarks of trading”, into which “no fiscal consideration or arrangement intruded itself in any way into any bargain that was made” and there was “merely an acknowledged reason which inspired one party” to enter into a trading transaction fell on the Harrison side of the line. However, Harrison did not mean that “a pure dividend-stripping transaction entered into with the sole object of making a fiscal profit without any view to a commercial profit was a trading transaction”. A transaction would fall on the Finsbury side of the line where:(1) it was “so affected or inspired by fiscal considerations that the shape and character of the transaction is no longer that of a trading transaction”;

(2) the fiscal arrangements were “inherently and structurally a part of the transactions which it was sought to describe as trading transactions”;

(3) “the fiscal element has so invaded the transaction itself that it is moulded and shaped by the fiscal elements.”

[409]He endorsed Megarry J’s approach of carrying out the analysis by looking at whether the elements of the transaction were explicable as having a trading purpose or as inserted “solely or mainly for the purpose of producing a fiscal benefit”. He concluded that Lupton fell on the Finsbury side of the line as it was one in which there was a “carefully worked out scheme which (in the hope of mutual financial benefit) was shaped and moulded by the fiscal possibilities” and that to describe the truly strange arrangements and the elaborate and unusual provisions as resulting in a share dealing transaction would be a “sheer delusion”.

[410]Overall, our view is that, in effect, Lord Morris also decided Lupton on the same basis as the other law lords by reference to the sole objective of the transaction as a tax recovery device. Each of their Lordships, therefore, approached the issue of determining whether the transaction was a trading one or not, by an objective assessment of the arrangements. On the facts of that case, it was because that objective assessment revealed that the object or purpose (or possibly the main object or purpose) was to recover tax rather than to deal in shares for commercial purposes, that the transaction was to be viewed as a “raid on the revenue” or a “tax recovery device” rather than a trading transaction entered into for commercial purposes.

[411]Where Lord Morris differed from the majority was in seeming to take a more restrictive view of the circumstances to be considered in making that objective assessment. It was only if that objective was apparent from the elements of the transaction itself, on the basis that tax considerations were somehow embedded in the transaction and had shaped or moulded it in some way, that it was dividend stripping and not a share dealing. Whilst that proposition of itself is not inconsistent with the approach of the other Lords, Lord Morris seemed to apply the test he put forward more narrowly in justifying a difference between Harrison and Finsbury on the basis of whether the tax recovery was reflected in the terms of the transaction, in particular, as regards the sharing of the benefit with the seller.

[412]As regards “equivocal cases” we note that Lord Simon said that if “the appearance of the transaction leaves the matter in doubt”, an examination of its “paramount object” had to be made. But again, that was a question of objective assessment as to what was the paramount object. This is clear from his conclusion that even if the transaction were equivocal its “paramount object” was as a tax recovery device, as he noted that it would be loss making but for the tax recovery.

[413]Whilst Lord Morris (and Lord Guest who agreed with him) could be said to be taking a narrower approach than the other law lords, we do not understand him to be laying down a principle that it is necessarily only where the fiscal arrangements are within the documents which effect the transaction that the transaction may be said to be so inspired by fiscal considerations or invaded by fiscal elements that it is no longer a trading transaction. He concluded that was the case on the facts of Lupton according to the terms of the share sale agreement. But, as he noted in Lupton and in Gurneville, each case must be examined on its own facts.

[414]In Gurneville again, Viscount Dilhorne, Lord Donovan and Lord Simon took the “broad” approach that it was necessary to look at whole picture in deciding, as Viscount Dilhorne said, that the transaction was “one designed, intended and carried out ... mainly to provide a basis for claims against the Revenue” or, as Lord Donovan said “for the sole or main purpose of dividend stripping” where “predominantly its aim was to exploit certain features of the existing fiscal system” or, as Lord Simon said, a “mere device to secure a fiscal advantage”.

[415]We note that, in Gurneville, the fact that, unlike in Lupton, the parties had engineered the pricing so that the taxpayer made a profit, which was not an insubstantial amount albeit much smaller than the expected tax recovery, did not affect the conclusion. Their Lordships did not consider that this affected the analysis given it was merely engineered as part of the tax scheme. Lord Donovan noted, for example, that the profit in that case was merely “part and parcel of the scheme” and Lord Simon that it was “merely a by-product of, or a disguise for, what was really a tax recovery device”. They and Viscount Dilhorne all noted that the profit was small compared with the tax benefit.

[416]Lord Morris (with whom Lord Guest again agreed) took a similar line as in Lupton in his comments that the fiscal matters were part of “the pith and substance of the transactions themselves” and, in effect, that the transaction did not bear “the stamp and mark of the trade of a dealer in shares” but rather was one where “its very structure and content reveals it as something different in kind”. However, again, we see nothing to suggest that the objective enquiry he advocated is necessarily limited to what is on the face of the documents implementing the scheme. Indeed, the reference to “structure and content” suggests that he also accepted that a broader enquiry was required.

[417]HMRC cited Coates v Arndale and Nova Securities as further support for their position, in particular, as it was held, similarly to Gurneville, that the making of a profit which was engineered to make the transaction appear commercial did not assist the taxpayer in the argument that the relevant asset was acquired for the purposes of its trade. As HMRC also noted, in both of those cases Lord Templeman cited passages from Lupton with approval.

[418]We note that the reason why the taxpayers in Nova Securities and Coates v Arndale did not succeed was not due to the fact that the transaction was fiscally motivated as a result of the intention to obtain the benefit provided by the appropriation rules of converting a capital loss into a trading loss (which could be group relieved). It is clear from Lord Templeman’s judgments that for a taxpayer to avail himself of the very benefit provided by the legislation is not something which can be held against him.

[419]Rather the issue was that, in order to qualify for the benefit provided by the rules, the taxpayer had to acquire the asset as trading stock. That required that an asset of the type the trader ordinarily dealt in was acquired for the purposes of the trade. It was that requirement which was not satisfied, as regards the property in Coates v Arndale and the shares in Nova Securities. The property was merely channelled from one investment company to another through a trading company which realised an engineered profit as a “timid veil to conceal that the lease was not being traded”. The shares were worthless such that they were not commercially saleable at any price and there was no commercial justification for their acquisition. In that context, as HMRC argued, Coates v Arndale, is relevant as demonstrating that an engineered profit will not necessarily mean that a transaction is regarded as commercial.

[420]BNP argued that the circumstances in both of those cases were very different to those in this case. The London branch clearly intended to trade with the dividend rights acquired and did so by selling the rights to an unconnected third party purchaser for a market value price. The resulting profit was not a mere book entry. There was a commercial justification for the difference in price paid by the London branch for the dividend rights and the price received by the London branch. We have commented on this in our conclusions on the facts below.

[421]BNP noted that, in any event, whilst purpose was clearly relevant to the specific issue in those trading stock cases, as regards whether an asset was acquired for the purposes of the trade with a view to resale, that is a different enquiry to the general one, in issue here, of whether a transaction is undertaken in the course of a trade. On that issue, BNP argued that Ensign Tankers establishes that fiscal motive is never relevant to whether a transaction is undertaken in the course of the trade.

[422]BNP submitted that is the effect of the passages in Lord Templeman’s decision where he criticised the decision in the lower courts and, in particular, the comments of Sir Nicolas Browne-Wilkinson V-C. The lower courts had to some extent focussed on the Lupton argument that the transaction was not undertaken as a trading activity because of the taxpayer’s fiscal aim to obtain allowances. Lord Templeman thought that approach was wrong; it would mean that all the beneficial fiscal consequences of a transaction could simply be ignored if the transaction had “fiscal motives as its paramount object” (as the Commissioners said) or was “essentially for the purpose of obtaining a fiscal advantage under the guise of a commercial transaction” as the Court of Appeal said. In his view, the approach in the Ramsay line of cases did not authorise the court to disregard all the fiscal consequences of a single composite transaction read as a whole because the transaction is a tax avoidance scheme.

[423]Whilst he saw the force of the Lupton argument because it was decided that tax avoidance was not trading, in his view, Lupton did not determine the Ensign Tankers case. On “the true analysis” of Lupton, “the trader did not trade at all” due to the lack of any actual financial loss, whereas, in Ensign Tankers, “on the true analysis” there was trading involving an expenditure of $3.25 million.

[424]In our view, in effect, he was saying that the Commissioners and the Court of Appeal had been asking themselves the wrong question by applying the approach that undertaking a scheme for tax purposes negated the entire tax benefit. It was in that context that he then made the remarks that he did not consider the Commissioners or the courts are “competent or obliged” to carry out the exercise advocated by Sir Nicolas Browne-Wilkinson V-C of assessing whether there was a sole object or paramount intention or weighing fiscal intentions against non-fiscal elements. Rather “the task of the commissioners is to find the facts and to apply the law. … The facts are undisputed and the law is clear.” In that regard, as set out in full at [388] above, he thought it plain that the taxpayer was carrying on a trading activity in which expenditure of $3.25 million of capital was incurred for the purpose of that trade as was required to generate a first year allowance in respect of that sum (noting that it was the purpose of the expenditure rather than the purpose of the transaction which was in point under the allowances rules).

[425]As noted, BNP interpreted Lord Templeman’s comments as meaning that fiscal motive is simply not relevant to assessing whether a transaction is undertaken in the course of a trade. It is not a case of weighing up or taking account of any tax avoidance motive. The sole test is to look at what taxpayer actually did, on an objective basis. When he noted, at the end of the passage set out in [388] above, that the Vice-Chancellor referred to authorities which sometimes illuminated and sometimes obscured the identification of a trading purpose but that “in every case actions speak louder than words” he meant, as illustrated in the authorities he then cited, that it is what is done rather than why it is done that matters. BNP asserted that there is a distinction between the trading stock cases, where an investigation into the purpose of the acquisition is specifically required by the legislation, and the situation in Ensign Tankers where the question was simply whether the partnership was trading. They asserted that Lord Templeman concluded that it was entirely irrelevant that the partnership “only engaged in the film trade for the fiscal purpose of obtaining a first year allowance” and noted that he said the transaction amounted to a trading transaction “whatever its design” (see [389] above).

[426]In BNP’s view this means that what it referred to as the third proposition in Lupton, that if a transaction is ambiguous or equivocal, then you can take into account tax avoidance purpose, is wrong. The key distinction, in BNP’s view, is as between the situation where(a) the taxpayer is dealing in an asset not dealt in by him in his trade at all, being in Lupton and Gurneville, an asset comprising a tax reclaim and

(b) the taxpayer deals in an asset of the sort he typically deals in albeit that the activity gives rise to fiscal benefits, such as the acquisition and exploitation of the film in Ensign Tankers leading to a capital allowances claim.

[427]BNP concluded that this is an Ensign type of case rather than a Lupton/ Gurneville type of case. The true nature and substance of the transaction entered into by the London branch is the acquisition and realisation of a financial asset, essentially a series of cash flows of a sort habitually dealt in by the London branch, in the course of its existing banking/financial trade. Like in Ensign, on an objective analysis, the London branch carried out the transaction in the course of its trade; the expenditure incurred by the London branch was “real and not magical”. The fact that BNP may have wanted to obtain a fiscal advantage does not alter the nature of the transaction.

[428]HMRC countered that Lord Templeman simply meant that there was no need in that particular case to look at paramount intention or object to establish whether there was a trading transaction. In that case, as he said, “the facts are undisputed and the law is clear.” They considered it is apparent from his conclusion that purpose was relevant when he identified as the real issue that “the expenditure of capital for the purpose of producing and exploiting a commercial film is a trading purpose.” In their view, it is implicit from the examples he gave of cases where intention sometimes illuminated and sometimes obscured the identification of a trading purpose, that he accepted that intention/purpose could be relevant. When he said “actions speak louder than words”, he was saying merely that was the best evidence of what was going on, to look at the objective features of the transaction, rather than at a person’s professed motivation.

[429]We note that HMRC’s interpretation that Lord Templeman simply meant that the exercise of deciding on sole or paramount intentions was not required in Ensign Tankers, is consistent with Lord Templeman’s view that it was very plain that the transaction in that case was a trading one. Given how plain he considered that to be, such an exercise would be unnecessary. That also accords with the statement that the Commissioners/courts were not obliged to carry out this exercise. However, this interpretation does not sit well with the statement that the courts/Commissioners are also not “competent” to carry out that task indicating this was intended to be a more general comment.

[430]Overall, we consider that Lord Templeman meant this as a general statement. In the context in which the comments were made, it appears his concern was to avoid the assessment of the effects of a transaction becoming subject to an exercise in weighing up fiscal and non-fiscal motives, on the basis that a sufficiently predominant fiscal intent would necessarily negate all its beneficial fiscal effects. To the extent, therefore, that there could be any argument that Lupton set out a special principle (such as Ramsay was once thought to do) that in any case where there is a predominant fiscal motive there is no trading transaction, we would say that Lord Templeman rejected that. It is a question in all cases of finding the facts and applying the law.

[431]The lower courts had gone awry, so he appeared to suggest, in simply not analysing the true effect of the transaction but rather in assuming that, if there was a predominant fiscal intent (on the weighing approach), that negates trading and thereby, in those particular circumstances, the whole of the first year allowances claimed. However, in Lord Templeman’s view, as properly analysed, that did not accord with the factual and commercial reality. On the particular facts, it was not the case that there was no trading transaction at all, as in Lupton, but that there was a plainly a trade in which excessive allowances were claimed.

[432]Lord Templeman noted, as in the disputed passage, that the task of the courts is to construe documents and analyse facts and to ensure the taxpayer pays the amount of tax which is consistent with the true effect in law of the taxpayer’s activities. In other words, it is a question of the correct analysis of the facts and law and not simply that, if there is a predominant fiscal motive, it must follow there is no trade (albeit that was the result in Lupton).

[433]It does not follow from this that Lord Templeman considered that the object or purpose of a transaction, whether fiscal or otherwise, is irrelevant to the fact finding exercise of whether a transaction is undertaken in the course of a trade. Indeed, we consider that it is implicit in his acceptance of Lupton, as a case where tax avoidance was held not to be trading, that he accepted that the object of the transaction may be relevant. His comment that “actions speak louder than words and the law must be applied to the facts” indicates that the facts, including as to the object or purpose of a transaction, are to be found on an objective assessment of what people do rather than their professed motivation for what they do. In accordance with the approach in Lupton, object or purpose is to be discerned from an objective assessment.

[434]On any view, it is clear that Lord Templeman did not take any issue with the decision in Lupton which would indicate that it is not good law. At the most, he rejected it as authority for the proposition that there is a special principle that the commercial effects of a transaction can be defeated in their entirety by reference to fiscal motive and/or that any enquiry as to the object of a transaction is to be made by a subjective enquiry into the taxpayer’s mind.

[435]We do not see that the distinction, which BNP drew, based on whether the taxpayer is dealing in the sort of asset he typically deals in takes the debate any further forward. On a narrow component by component approach, as in Lupton and Gurneville, the London branch implemented a transaction which had the apparent features of its usual trade of dealing in financial assets. However, as in those cases, it does not necessarily follow that, on wider examination of all the circumstances, it was in fact undertaken as part of that trade.

[436]We do not see any support in Ensign Tankers itself for the proposition that this transaction falls, as BNP sees it, on the Ensign side of the line as a trading transaction albeit with fiscal benefits. The effect of the House of Lords decision was to permit the taxpayer to obtain the benefit of the intended fiscal consequences of the transaction to the extent that they accorded with the commercial and economic reality which was that the partnership had spent $3.25 million for the purposes of its film trade. Accordingly, the taxpayer was able to claim capital allowances on $3.25 million but not on the additional amount it claimed it was also entitled to allowances on, as it had not actually incurred that cost (as a result of the non-recourse loan arrangements). Such a conclusion does not support the contention that a transaction under which the whole of the expenditure in question was incurred with the effect of generating a tax loss, with no actual financial loss, is a trading transaction. Rather, on the face of it, it falls within the description Lord Templeman gave to Lupton as a case where, viewing the transaction as a whole, the taxpayer had made no loss at all such that, on a true analysis, there was no trade at all.

[437]BNP argued that the decision of the Court of Appeal in New Angel Court supports its view that tax motive is no longer relevant to the trading issue. This was based on Jonathan Parker LJ’s comments that Lord Templeman’s remarks in Ensign Tankers mean that, in assessing whether a transaction is a trading one, “the relevance of fiscal purpose and, in particular, the distinction which [Sir Nicolas Browne-Wilkinson V-C] sought to drew between sole and paramount purpose are no longer good law.”

[438]HMRC countered that New Angel Court was merely a case (like Coates v Arndale and Nova Securities) where the parties were seeking to exploit the “mini code” relating to the appropriation of items as trading stock. Fiscal considerations clearly played a part in the decision to transfer the properties to a dealing company, but the fiscal considerations were ones permitted under the legislation itself. Essentially, the Court of Appeal held that, for the purposes of the “mini code”, the focus is on the reasons for the acquisition and the purposes of the acquisition which, in that case, were for resale at a profit. The dicta that the presence of fiscal considerations does not alter the analysis should be read solely in the context of the fiscal considerations which were authorised by that “mini code”. It was not necessary to Jonathan Parker LJ’s decision to form any view on the continued applicability of looking at intentions in equivocal cases, given that Lupton and Ensign were not concerned with the “mini code” but with the general meaning of “trading”.

[439]New Angel Court does not, in HMRC’s view, provide any real guidance on the relevance of the type of fiscal considerations at play here and does not undermine the relevance of intention or purpose in equivocal transactions. In any event, they consider that this is not an equivocal transaction. This is a clear case where fiscal considerations have distorted the transaction to the point where it was not a trading transaction but rather a tax-saving exercise. BNP responded that Jonathan Parker LJ’s comments were not confined to the “mini code”. Whilst that was the ultimate focus of the decision, these conclusions were expressed as part of a general discussion of the relevance of fiscal motive and were based on cases which do not concern the “mini-code”.

[440]It seems to us that, as HMRC argued, the decision in New Angel Court accords with the earlier ones of Coates v Arndale and Nova Securities. As Jonathan Parker LJ makes very clear, a transaction involving the appropriation of an asset to trading stock cannot be defeated purely because (as must inevitably be the case) the motivation is a fiscal one to obtain the very benefit for which the legislation provides. The comment about the effect of Lord Templeman’s judgment in Ensign Tankers was made in the context of more general comments about cases involving a consideration of the role of fiscal purpose more widely. Whilst we would not necessarily disagree with the statement that Lord Templeman’s comments mean that the distinction the Court of Appeal sought to draw between sole object and paramount object is not good law, we agree with HMRC that this observation was not necessary for the decision in New Angel Court. It does not form part of the binding precedent set by that case. In any event and as we have set out above, we do not consider that the relevant comments in Ensign Tankers mean that object or purpose is not relevant to the factual assessment of whether there is a trading transaction or not.

[441]HMRC referred also to a recent decision by the tribunal in Clavis Liberty 1 LP (acting through Mr D J Cowen) v Revenue and Customs Commissioners

[2016] UKFTT 253 (TC)

that arrangements similar to those in this case were not undertaken as trading transactions. There is little discussion of the issue but the tribunal essentially applied a Lupton approach noting that there was nothing in Ensign Tankers or New Angel Court which precluded them from doing so. We agree with the appellant, however, that the brevity of the tribunal’s discussion of Lupton in Clavis Liberty means that it does not advance the arguments as to the correct analysis of Lupton or the cases related to it.

[442]Finally, HMRC said that the general approach of the tribunals and courts on the question of trading should be considered and, in particular, the two observations of the Upper Tribunal in Samarkand being, at [86(2)], the fact that the question of trading requires a close analysis of what actually occurred and, at [87], the need to look at matters as a whole. BNP agreed that, in making the required objective assessment, it is necessary to view what the taxpayer did in its proper context. As set out in BNP’s contentions on the facts below, however, there is nothing in the context of this transaction which detracts from it being a trading transaction.

[443]In conclusion, we see nothing in Ensign Tankers or New Angel Court which detracts from the approach set out in Lupton and Gurneville. Those cases do not establish a special principle that the presence of a tax avoidance purpose negates the commercial effects of a transaction. Rather, in line with the modern approach to statutory interpretation in a tax avoidance context, we must make an assessment objectively, on the basis of the available evidence, whether in all the circumstances of the cases, the transaction is undertaken in the course of a trade. This would not be the case if that assessment reveals that, in reality, the transaction is merely a tax repayment device.

[444]Following the approach in Lupton and as stated in the Upper Tribunal case in Samarkand, at [86(2)], the question “is not to be answered by asking whether the transaction is of a type which may in other cases have been held to constitute trading … but by examining the particular transaction in question”. However, as the Upper Tribunal also went on to say in Samarkand, at [87], this does not mean that a narrow view should be taken. Rather the transaction should be examined fully in its context, as they noted was supported by the approach taken by the Court of Appeal in Eclipse Film Partners No 35 LLP v HMRC

[2015] EWCA Civ 95

, in referring at [111] to:

“‘an unblinkered approach to the analysis of the facts’, a ‘realistic approach to the transaction’ and to it being ‘necessary to stand back and look at the whole picture and, having particular regard to what the taxpayer actually did, ask whether it constituted a trade’.”

Conclusions on the facts

[445]In line with its legal arguments, BNP said that the focus must be on the shape and character of the transaction undertaken by the London branch, which in its view, was plainly a trading transaction as set out above. BNP emphasised that, on an individual basis, the London branch realised a profit (of £402,372 (after costs)) which, as we accept, was computed so that it was not out of kilter with what a trader would expect to receive on a trade in this type of financial instrument. BNP accepted that the transaction must be considered in its context but, in its view, there is nothing in that wider background which detracts from this conclusion. In particular, BNP asserted that all of the other BNP group entities involved in the transaction acted commercially; they all realised a pre-tax profit and, taking into account the s 730 benefit, a post-tax profit. The overall before-tax profit was £1,134,205 (before taking account of costs on the transaction).

[446]As regards the group position, both Mr Peters and Mr Demon emphasised that the arrangements provided funding for the BNP group at an attractive rate, lower than the market rate, giving rise to a pre-tax profit on an overall basis. Both also maintained to some extent that, whilst obtaining the s 730 advantage was an important part of the transaction, the BNP group also had the purpose of obtaining this “attractive” financing. Mr Demon said in his witness statement that the transaction was ultimately approved by the TCC because it provided this funding at a low rate. Whilst the internal email correspondence focuses on the tax purposes, other documents, such as those for presentation at the approvals meetings, give obtaining cheap funding as the rationale.

[447]HMRC accepted that, looking at the transaction narrowly, the analysis is that the London branch bought and sold an asset which it generally buys and sells in the course of its trade. However, looking at the wider picture and, in particular, the steps taken to enable the transaction to take place, it was wholly uncommercial. Neither the dividends, nor a market for them, existed in the real world and, therefore, one had to be created on a bespoke basis. Whilst the sale to ALIL was to an unconnected third party, the manner in which BNP acquired the right to the dividends from BNP Lux (and instigated its creation) was wholly uncommercial. Tax essentially infected every stage of the arrangements and analysis.

[448]We accept that the fact that the dividend rights were created under a bespoke internal arrangement does not, of itself, render the transaction a non-trading one. The dividend rights created were analogous to other financial instruments in which the London branch dealt. On the face of it, the transaction contained the elements, namely, the purchase and sale of a right to an asset of a type the London branch typically dealt in for a resulting profit, which are required for it to fall within the course of its trade. However, as advocated in Lupton itself and in recent cases such as Samarkand, looking at all of the circumstances, including the inception of the transaction and manner of its implementation, we find that this transaction was not trading in financial instruments for a profit but was a tax recovery device.

[449]The transaction had certain effects which would usually be regarded as indicating the commerciality of a transaction in that funding was obtained from a third party, ALIL, and the London branch and the overall BNP group made a profit on a before-tax basis (and on an after-tax basis taking into account the s 730 benefit). However, it is clear that these effects were designed as an integral part of the scheme purely with a view to achieving the objective of realising the s 730 benefit. As in Gurneville and Coates v Arndale, the “commercial” effects were “part and parcel of” the scheme and no more than a “timid veil” to give the transaction the “verisimilitude of commerciality”. The fact that, due to the nature of the planning, the effects resulted from more elaborate arrangements than in those cases, involving many group members, does not detract from this. In each case, the effects were not ends in themselves but rather engineered “commercial” incentives considered necessary for the planning to succeed. It is clear from the design of the scheme and the evidence of the witnesses that every element was implemented with the single goal of achieving the desired tax benefit.

[450]Moreover, the reality is that the creation of the dividend rights and the generation of a before-tax profit for BNP London and overall for the group was made possible only through putting in place what were, but for the obtaining of the s 730 benefit, inherently uncommercial arrangements. Both Mr Peters and Mr Demon accepted that the transaction did not make sense and would not have been entered into without the s 730 benefit. That is, in any event, apparent from the structure and economics of the arrangements.

[451]It is clear that this was not a case of BNP seeking funding of £150 million which it then structured to optimise the tax effect. On the contrary, OF adapted a proposal which was designed by Swiss Re and Dresdner to obtain a s 730 benefit to include commercial elements, such as the “commercial” loan from ALIL, which were considered necessary to achieve the benefit and to gain the required internal approvals. As Mr Peters said a number of times and, as is clear from the internal correspondence, the view was that each party needed to be “commercially incentivised” to enter into the arrangements; otherwise the transaction would simply not get done. The original proposal did not contain these commercial elements but OF set about structuring them into the scheme.

[452]The London branch could not have obtained the s 730 benefit under this structure if it had invested in the shares in HIL itself and sold the dividend rights. There needed to be a prior sale of the dividend rights to the London branch from an offshore entity located in a favourable tax jurisdiction so that it did not suffer any UK tax charge on the sale. Hence, having come across a problem with the use of an Irish subsidiary, BNP Lux was selected as the relevant party. Mr Peters accepted that there was no reason for the inclusion of BNP Lux in this role except to be able to take advantage of the perceived effect of s 730.

[453]It was identified in the early stages that it was crucial to the success of the scheme that the London branch made a profit from the purchase and onward sale of the dividend rights. This meant that there had to be a justifiable difference in the price which the London branch paid for the dividend rights and that which it received from ALIL of a sufficient amount to generate a profit of the kind a dealer in this type of asset may expect to receive. It was for this reason that OF decided that the “loan” transaction should take place with a UK tax payer entity as the counterparty. The rationale for the price differential was that, as such an entity would not be taxable on the dividends received, it would be prepared to pay a higher price for the dividend rights than the London branch, which would be taxable on the receipts. The justification is somewhat notional given that the plan was that the London branch would not actually be in receipt of the dividend stream.

[454]The oddity in the carefully “commercially” constructed pricing arrangements stems from the fact that, in order to create a right to a principal repayment of £150 million (carrying a fixed rate dividend of the required amount), BNP Lux had to subscribe £210 million for the shares in HIL. The additional amount of £60 million was required to ensure that HIL had sufficient funds to generate enough income to pay the fixed rate dividends and to fund its corporation tax liability on interest income received on the funds it deposited with the Dublin branch given that the dividends were not tax deductible. However, as BNP Lux had to receive less than £150 million on the sale of the dividend rights (in order to generate a profit for the London branch on the onward sale to ALIL) and indeed as it could not realistically expect to receive any higher amount for a right of that nature, the creation of the dividend rights and their onward sale was inherently loss making for BNP Lux due to its ongoing funding cost on the remainder of its £60 million investment.

[455]The financial model which Mr Bayfield sent to Ms Majchrzak Gilot on 12 December 2005 showed that, following the sale of the dividend rights, as at 31 December 2005, it would have a loss of £1,184,000. Mr Stanton calculated that for BNP Lux to make a profit over the three year term of the deal with ALIL, it would need to receive a price of £157 million. He also calculated that had the transaction run its course of three years, the loss would have been in the region of £9 million on a before-tax basis or around £6.4 million on an after-tax basis. This loss or cost is recognised in the papers presented by GF and OF to the TCC and in the later GF memo circulated on 10 January 2006 (see [229] above). Mr Demon agreed that the loss of over £6 million identified in that memo was essentially a funding cost on £60 million or that it arose because the £150 million of funding “was not tax deductible” which he said was the same thing.

[456]As a standalone proposition, therefore, it was inherently uncommercial for BNP Lux to fund the investment in HIL to the tune of £210 million. It was acknowledged by the witnesses that BNP Lux would not have created and capitalised HIL but for the fact it was presented with the sale of the dividend rights and the “rump” as a composite package to incentivise it. From the early stages, it was proposed that BNP Lux would be able to sell the shares in HIL at later stage for an amount which would meet its funding costs and ensure that overall it would make a profit. Although the UK tax team insisted that this could not take place otherwise than by way of option arrangements with exercise only after the end of the tax year and the identity of BNP UK as the acquirer was not confirmed for some time, it is clear that there was a settled plan from the outset. Somewhat at odds with the lengths that BNP went to at the time to avoid the perception of the sale as an inevitability, BNP argued that the fact that it was always envisaged there should be such a sale means that the arrangements were not uncommercial from BNP Lux’ perspective. It was always the intention that BNP Lux would be suitably rewarded for its role in the scheme.

[457]We accept that it was intended from the outset that BNP Lux would be made whole as regards its funding costs and receive a profit through the later sale of the “rump”. However, the fact that it was part of the design of the scheme that BNP Lux was “commercially incentivised” in this way does not render the scheme any the less artificial or uncommercial overall. We note that BNP Lux was fully compensated and received a profit only by dint of the fact that another group member, BNP UK, paid a substantial overvalue for the “rump” with funding provided by BNP. BNP UK recognised an impairment loss due to the overpayment of £2.7 million. BNP had an on-going funding cost on £62.7 million of which £59 million accrued interest at a floating rate which was effectively swapped into fixed rate interest under the swap with BNP Dublin and the remaining balance of £3.7 million probably accrued interest at a floating rate. In effect, therefore, the loss making situation was simply transferred to other group members.

[458]Mr Peters said that BNP did not suffer any real loss as regards the overvalue paid for the shares in HIL of £2.7 million; all that was happening was that its investment of that amount in BNP UK moved to another of its subsidiaries, BNP Lux. We do not follow that given that, aside from the profit element, the balance of the overpayment went to payment of an expense for the group, BNP Lux’ funding costs. As regards the on-going funding costs for BNP on £62.7 million, the appellant came back to its argument that, notwithstanding that this resulted in an individual financing cost for BNP, the overall position for the group was a profit of £1.13 million on a pre-tax basis (before taking account of costs).

[459]BNP placed some emphasis on the fact that the price paid by ALIL to the London branch and by it to BNP Lux can be said to reflect an “arm’s length” price which reflected the fair value of the dividend rights. However, in our view, this does not provide any assistance to the argument that this was a trading transaction bearing in mind how the price to ALIL was actually set and how the pricing of the other elements worked.

[460]It is clear that the size of the transaction of £150 million was determined by reference to the tax capacity available to the London branch. A&L clearly had an appetite for a higher investment. They expressed an interest in investing up to £450 million but the TCC determined the transaction size as £150 million in view of the tax capacity. So, whilst the price of £150 million paid by ALIL was “arm’s length” in that it was subject to agreement with a third party counterparty and that party’s appetite for investment, it was in fact capped at the size of the tax benefit which the BNP group wished to obtain.

[461]The other pricing and profit levels were worked out backwards from that fixed starting point of £150 million to give what the London branch thought was an acceptable level of pre-tax return for each entity involved in the transaction whilst ensuring ALIL achieved its expected rate of return. None of the pricing elements of the transaction involved any actual valuation of the relevant assets in question. A “goal seek” method of calculation was used to calculate the price paid by the London branch to BNP Lux for the dividend rights and the price paid by BNP UK for the purchase of the shares in HIL from BNP Lux. This involved plugging into the computation the desired figures for the profit to be achieved by ALIL and BNP Lux and other known amounts, such as the funding costs, to give a resulting price. The same methodology could have been used whatever the starting point of the transaction size.

[462]As the price to be paid by ALIL was fixed, the price to be charged to BNP Lux had to be lower to give the London branch its desired rate of return, as an amount which the London branch would expect to receive on a transaction of this type. Our conclusion from all the evidence is that the OF London team were indifferent to the level of profit for the London branch or BNP Lux other than ensuring that it was sufficient to incentivise each party to enter into the transaction. Ms Majchrzak Gilot made it plain she wished to obtain a particular level of profit for BNP Lux, in return for its participation in the plan, which was initially stated to be £500,000. When and why this increased to £750,000 is not readily apparent from the documents. However, given Ms Gilot’s approach, it is not unreasonable to suppose that it was in response to BNP Lux’ desire to be appropriately compensated for taking part in the arrangements. 463.In that context, a debate over the appropriate price for the sale of the dividend rights by BNP Lux to the London branch is somewhat redundant. It is not the price paid for the dividend right as such which was uncommercial. Once the price or transaction size was fixed at a particular level with the third party, being £150 million, that, in turn, fixed the parameters of the price achievable by BNP Lux. As noted, the oddity arises because BNP Lux had to invest £210 million to generate the dividend rights which, of necessity, under the parameters set by those rights themselves, it could only sell for £150 million at the most. Having divested itself of the asset for £60 million less than it paid for it and deprived itself of the income from the asset, BNP Lux was in a substantial loss making position. The reason it did this was in order to subsidise HIL which would not otherwise be able to meet its obligations due to the non-deductibility of the fixed rate dividends.

[464]HMRC argued that, leaving aside the s 730 benefit, there was no commercial rationale for borrowing in the manner that BNP did. It borrowed more expensively compared with the rate at which it could have borrowed conventionally as well as incurring much larger legal costs. BNP were limited in what they could do with the “loaned” monies due to the termination provisions which meant that the funds could be deposited only at 1 month Libor. It was only the perceived success of this scheme that enabled BNP to enter into a transaction that it would not normally have done and which was forecast to cause post-tax losses of over £ 6m. In HMRC’s view, the “sleight of hand” was to ignore these losses and simply lay claim to a pre-tax figure of beating the market by 38.6 basis points. However, if a pre-tax equivalent figure had been used instead, this would have shown the “loan” element of this scheme for what it was; loss making. In their view, that there was no real interest in the funding is shown by the fact that the transaction size was capped at £150 million by reference to the tax capacity despite ALIL’s willingness to invest more and that, due to the change in the law, there were no further transactions of this nature. 465.BNP accepted that, without the s 730 benefit, the borrowing would have been uncommercial but asserted that the same is true of any borrowing: without a deduction for the interest paid the borrower’s post-tax position is uncommercial. BNP noted that, taking account of the s 730 benefit, the borrowing was commercial from a post-tax point of view. In any event, in BNP’s view, the post-tax effect of a transaction is to be derived from its characterisation, not the other way around. 466.BNP argued that the fact that the funding could be terminated on short notice and that upfront costs were incurred does not affect the analysis. Traders commonly incur costs without knowing whether they will generate profits: this sort of speculation is not inconsistent with trading. The money “lent” to the BNP group, generated real benefits for the group and was, in Mr Stanton’s words, “good money to have”, albeit that its potentially short-term nature limited its utility. They noted that it was in the interests of the A&L group to allow the transaction to run its three-year term and that BNP acknowledged that the right to early termination would not be exercised because of the risk of damaging the commercial relationship with the A&L group. In BNP’s view, therefore, on a realistic view of the facts, neither side was likely to exercise those termination rights.

[467]HMRC regarded BNP’s argument as wholly circular. On their view, the arrangement only makes sense post-tax if the s 730 benefit is available. However, that benefit is only available if BNP is acting commercially and, it is only acting commercially, if it can take that benefit into account. Also, on BNP’s analysis, the benefit is only available if the narrow transaction is a trading transaction as opposed to the commerciality of the borrowing arrangements as a whole.

[468]Overall, each entity in the group realised a profit on a pre-tax basis which, in aggregate, was £1,134,205 (before deducting costs relating to the transaction). This arose due to the difference in the fixed rate interest payable by the group on the “loan” of £150 million at 4.354% and that received by the group under the swap arrangements of 4.74%. The “loan” from ALIL was, therefore, on the face of it at a rate which represented interest at 0.386% (or 38 basis points) below 1 month Libor. It was not disputed, however, that due to the absence of a tax deduction for the fixed rate dividends the transaction would have been loss making on an after-tax basis but for the obtaining of the s 730 benefit. The headline rate of borrowing, absent the s 730 benefit, would in fact have been expensive at Libor plus 1.47%.

[469]We find it difficult to see that in circumstances where, absent the s 730 benefit, the funding obtained would be very expensive and the arrangements required to fund obtaining the loan wholly uncommercial as set out above, it can be said that obtaining funding at a cheap rate was in any sense an objective of the transaction. Mr Demon and Mr Peters acknowledged that the transaction was also designed to obtain the s 730 benefit but, in their view, that simply meant this “cheap funding” was obtained in a highly tax efficient way. They said it was no different to a plan vanilla loan, it was just that the hoped for tax deduction was much larger.

[470]However, the fact that the financing was wholly uneconomic in the absence of the s730 benefit of itself indicates that, on the contrary, the objective was solely to obtain the s 730 benefit, with the obtaining of financing at an otherwise expensive rate being a necessary element in achieving that objective. Overall, as set out above, every element of the design of this scheme indicates that this was not a transaction undertaken to obtain “cheap” financing which was simply structured in a tax efficient way. On the contrary, it was an expensive financing obtained specifically to generate the s 730 benefit.

[471]We also note that, despite Mr Demon’s assertions that obtaining “cheap” funding was a consideration for the group in entering into the transaction, there is no indication that that was the case from the internal correspondence or the minutes of the meeting of the TCC approving the transaction. On the contrary, this indicates that the sole motivator was the obtaining of the s 730 benefit (as regards the correspondence, see in particular [54], [71] and [197]). All indications are that the references in the approvals documents to obtaining the cheap financing as an objective are merely window dressing; Mr Peters accepted that the group wanted to present the transaction in the best light and the relevant statements were somewhat “over-egging” matters. In any event, as noted, we consider that the fact that the sole objective was to obtain the s 730 benefit is inherent in the very economics and structure of the transaction.

[472]We note that Mr Peters set out calculations in which he sought to explain the “cheap” financing obtained by the BNP group under the loan from ALIL on the basis that, in effect, the group was sharing in the benefit to ALIL of receiving non-taxable income on the loan in the form of the fixed rate dividends. He said that this was analogous to how a preference share lending or, as he considered was a closer analogy, a deposit surrogate transaction, would be priced. There was some debate with Mr Stanton on this but we do not consider that whether the transaction can be said to fall in one or other of these categories adds anything to the issue. Nor can we see any substantiation from these calculations for the assertion that this was a financing in which the BNP group were sharing ALIL’s tax benefit. The calculations take into account the assumed tax position for the A&L group on the loan of £150 million but simply ignore the tax effect for the BNP group. The fact remains that, absent the s 730 benefit, the loan was expensive due to the lack of a deduction for the fixed rate dividends. Mr Peters sought to justify this on the basis that, as he said was the case in a deposit surrogate transaction, a group needed to have a deduction at least equal to the non-deductible dividends which was the case (and much more so) due to the s 730 benefit. Again, this comes back to the commercial justification for the transaction being entirely dependent on the obtaining of the s 730 benefit.

[473]We note that the documentary evidence shows that in fact those involved in Project Sumatra regarded this as an expensive financing, the cost of which, due to the non-deductibility of the fixed rate dividends, reduced the value of the potential s 730 benefit the longer that it went on. The papers prepared by GF, OF and UK tax for the purposes of the approvals process all acknowledge that the tax benefit was obtained upfront and the GF and OF papers suggest that the subsequent cost of the non-deductible fixed rate dividends was seen as eroding this benefit. The email from Mr Bayfield on 19 November 2005 (see [106] above) was explicit that the best thing for the group was to terminate the transaction immediately after it was implemented because “we will have secured our tax advantage, and therefore paying [dividends] to [A&L] at the rate agreed is a drag on the group”. It is clear from the correspondence and it was accepted by Mr Peters that the only reason the BNP group did not in fact terminate straight away was because of the relationship with A&L. We also note that matters were structured so that, if there was an early termination, ALIL would not suffer any material cost; that was the reason for the arrangement regarding ALIL’s legal costs and for the provisions in the swaps resulting in no material breakage costs.

[474]Ultimately, this transaction only made any commercial sense for the London branch and the BNP group if the s 730 benefit was obtained; it was wholly dependent upon that benefit. Whilst matters were structured so that the London branch made a profit, it was generated only as a result of the wholly uncommercial investment by BNP Lux of an excessive amount of £210 million to generate a right to a principal of £150 million plus dividends which left it in a loss-making position. BNP Lux was in turn subsidised by BNP funding BNP UK to buy the shares from it at an overvalue so that BNP Lux recouped its funding costs and received a profit of the required amount. BNP itself was in effect compensated for its loss on the on-going funding cost it took over from BNP Lux only if the s 730 benefit was received. Whilst there was, overall, a small pre-tax profit for the BNP group, this was entirely wiped out by the fact that the fixed rate dividends were not tax deductible. The transaction only came good on a group basis if the s 730 benefit was obtained.

[475]We agree with HMRC that the argument that the obtaining of the s 730 benefit itself renders the arrangement commercial, as more than covering the lack of a deduction for the fixed rate dividends, is wholly circular. As in Lupton and Gurneville, on an objective assessment, the transaction was designed, intended and implemented as a tax recovery device and not as part of the London branch’s financial trade. This is a case where the transaction is simply a tax recovery device for the obtaining of a tax loss where there is no economic loss. In those circumstances, it can hardly be said, as it was not in those cases, that the commercial justification for the transaction is the very fact that it is designed to obtain the tax recovery. Deduction for expenditure Overview

[476]Our conclusion on the trading issue determines this appeal but, in case we are wrong on that issue, we have considered the question of whether the London branch is entitled to a deduction for the price paid for the right to the dividends in computing its trading profits under s 74(1)(a) ICTA. This provides that “no sum shall be deducted in respect of - (a) any disbursements or expenses, not being money wholly and exclusively laid out or expended for the purposes of the trade or profession”.

[477]We note that BNP argued that it is inherently improbable that, if a transaction is found to be undertaken in the course of a trade notwithstanding that it is motivated by tax considerations, those tax considerations would be held to suffice to deny a deduction for expenditure incurred under that transaction. However, whilst not perhaps likely, it is possible given that the two tests are somewhat different, in particular, as explained below, as the s 74 test is applied by reference to the subjective intentions of the trader.

[478]The meaning of this test is well established in the cases. The word “exclusively” means that if the expense was also incurred for some other non-trading purpose, it is not deductible. The “wholly and exclusively” issue is to be determined by the object of the taxpayer in incurring the expense. This is a question of fact but in making that factual assessment the tribunal must observe a number of principles. As summarised by Millet LJ in the case of Vodafone Cellular v Shaw

[1997] STC 734

(at 742 f to j) and explained by the Upper Tribunal in Scotts Atlantic Management Ltd v HMRC

[2015] UKUT 66 (TCC)

[2015] STC 13219

(‘Scotts Atlantic’), at [47] to [53], these are as follows:(1) The words “for the purposes of the trade”’ mean “to serve the purposes of the trade”. They do not mean “for the purposes of the taxpayer” but for “the purposes of the trade”, which is a different concept. A fortiori they do not mean “for the benefit of the taxpayer”.

(2) Save in obvious cases which speak for themselves, ascertaining the taxpayer’s object in making a payment involves an inquiry into the taxpayer’s subjective intentions at the time of the payment.

(3) The object of the taxpayer in making the payment must be distinguished from the effect of the payment. A payment may be made exclusively for the purposes of the trade even though it also secures a private benefit. This will be the case if the securing of the private benefit was not the object of the payment but merely a consequential and incidental effect of the payment. In Scotts Atlantic, the Upper Tribunal commented (at [52]) that:

“another way of phrasing this is that a merely incidental effect of expenditure is not necessarily an object of a taxpayer in making it. However, as Lord Brightman’s well-known example in Mallalieu (see [1983] STC 665 at 669, [1983] 2 AC 861 at 870) of the medical consultant going to the South of France to treat a friend shows, it may be the case that in fact what would be an incidental effect in some circumstances could be an independent object in others. What the FTT must not do is to conclude that merely because there was an effect, that effect was an object.” (4) Although the taxpayer’s subjective intentions are determinative, these are not limited to the conscious motives which were in his mind at the time of the payment. Some consequences are so inevitably and inextricably involved in the payment that, unless merely incidental, they must be taken to be a purpose for which the payment was made. As the Upper Tribunal said in Scotts Atlantic (at [53]), another way of putting that is that the tribunal must take “a robust approach to ascertaining the purposes of the taxpayer.”

[479]There is no dispute that, on the basis of the principles set out above, whether the amount paid by the London branch for the right to the dividends was incurred wholly and exclusively for the purposes of its trade is to be determined by ascertaining what its object was in making the payment, which is distinct from and is not necessarily determined by the effect of the payment. This requires an examination of the London branch’s subjective intentions.

[480]Again, the dispute centres around the impact on the analysis of the potential tax advantage which it was hoped the transaction would achieve. In outline, BNP’s position is that the London branch’s object in incurring expenditure on the right to the dividends was to acquire that right, which it acquired for its fair value, for sale to ALIL at a profit. That the London branch could deduct the cost of the dividend rights in computing its trading profits for tax purposes simply followed from this, as a natural consequence. The London branch merely had the hope and expectation that it would be able to sell the rights tax free. The tax effects of the transaction were, therefore, no more than incidental or consequential effects of the transaction.

[481]HMRC relied on Scotts Atlantic, as authority that, where a payment is made with the avoidance of tax as one of the objects, that is sufficient to disqualify the whole payment from being deductible; at least one object of the London branch in incurring expenditure on the acquisition of the right to the dividends was to avoid corporation tax. Each pound was expended as part of a scheme to reduce the London branch’s taxable income and generate a tax loss. This was not merely an incidental effect of otherwise qualifying trading expenditure; in reality, it was the only reason why the London branch acquired the right to the dividends and why the dividend rights existed in the first place. Therefore, this duality of purpose prevents the whole payment from being tax deductible.

[482]BNP argued that the cases of Kilmorie (Aldridge) Ltd. v Dickinson

[1974] 1 WLR 1594

(HL) (‘Kilmorie’) and Drummond v Revenue and Customs Commissioners

[2009] STC 2206

(‘Drummond’) in which the taxpayer was found to have a tax avoidance purpose, support its position and run counter to HMRC’s argument. In BNP’s view, those cases are clear authority that, where an amount is paid to acquire an asset as trading stock, which reflects the fair value of that asset, any tax benefit thereby obtained is an effect which is merely incidental or consequential upon the transaction. BNP regarded the case of Scotts Atlantic as clearly distinguishable from this case. Case law

[483]Kilmorie involved a scheme designed to avoid tax on the sale or the development of land by the Downes family. A company owned by the family entered into an agreement with a land owner giving it the right to develop land in return for a payment of £67,500 (to be paid by instalments as it sold leases of the houses built by it). Under the scheme, following a number of transfers, the rights under the agreement were transferred from Opendy (a company owned by the devisers of the scheme) to Kilmorie (a Downes family company), for £77,250 payable in instalments. It was clear that this was a gross overvaluation given nothing material had changed since the agreement had previously been transferred for £2,250. Essentially Kilmorie had to pay such a large sum so that £60,000 could reach the Downes trustees free of tax under the tax avoidance scheme. Kilmorie claimed that the first instalment of the price was deductible from its trading profits as money wholly and exclusively laid out or expended for the purposes of its trade.

[484]The House of Lords upheld the decision of the Special Commissioners that, as there was a dual purpose in making the payment, of which only one purpose was trading and the other was tax avoidance, none of the expenditure qualified as deductible trading expenditure. As much of the discussion centred on the Commissioners’ finding of fact we have set out the relevant passage (as quoted by the House of Lords) in full. In their view, the agreement:

“… was an essential prerequisite to the carrying out by Kilmorie of the development of the estate. It proved, moreover, in the event to be very much to the advantage of Kilmorie to enter into the first-mentioned agreement ……. We have, however, to consider the position at the time when the [agreement] was made, and against the background of the series of transactions which led up to it. So approaching the matter … we are of opinion that the [agreement] was entered into by Kilmorie with the objects both of enabling that company to develop the Landywood Estate and of facilitating the scheme for avoiding liability to income tax … In our view the latter object was on the facts of the case one of the main purposes, and not a mere secondary consequence, of the entering into by Kilmorie of the agreement, and the outlay totalling £19,240 was thus incurred by Kilmorie for dual purposes being purposes one of which was, and one of which was not, a trading purpose.”

[485]Lord Reid considered that the Special Commissioners “plainly meant” that Kilmorie would not have paid so large a sum to Opendy “but for their non-trading purpose of enabling the tax avoidance scheme to succeed”. He noted that neither party was acting as a free agent in its own interest (as Opendy was a subsidiary of the devisers of the scheme and Kilmorie was a Downes family company) and “both had been procured to play their part in the scheme”. He continued similarly to link the overinflated price with the tax avoidance scheme in saying that “the price was dictated by the scheme, and plainly had nothing to do with the market value of the rights sold”. He also thought it quite obvious that none of the companies acted in their own interests; they just did what Mr Downes and the devisers of the scheme wanted. He said that if a trader only had commercial motives the Revenue cannot merely say that he has paid too much; he may have been foolish or he may have had what could fairly be regarded as a good commercial reason for paying too much but:

“if it is proved that some non-commercial reason caused the trader to pay more than he otherwise would have done, then it seems to me quite clear that the payment can no longer be held to have been wholly and exclusively expended for the purposes of the trade. No authority is needed for so obvious a proposition.”

[486]Lord Reid is clear that it is only if there is “some non-commercial reason” which causes the overpayment that the payment fails the wholly and exclusively test. In the context of the findings of the Special Commissioners, which Lord Reid was considering, the non-commercial reason for the overpayment can only be the facilitation of the tax avoidance scheme.

[487]He concluded by noting that the statutory provision could well be read as meaning that, if it could be shown that a part of the expenditure was in fact wholly and exclusively for trading purposes, then that part was deductible. But he did not have to decide that because the Revenue had agreed that in this case £2,250 of the £77,250 paid would be allowed as a deduction being the then market value of the rights.

[488]Lord Morris based his decision on the tax avoidance motive (at 1609 B to D). He noted that Opendy only came into the story in order to facilitate the scheme and the amount of £77,250 had to be paid, not wholly and exclusively for the purpose of developing the Landywood Estate, but “mainly in furtherance of and in order to facilitate the scheme for avoiding liability to income tax”. He noted that, as the agreement did have some value, the disallowance of the whole of the £19,240 “bore somewhat hardly”. But he nevertheless considered that the appeal failed.

[489]Lord Wilberforce said, at 1616 B to D, that what the Commissioners meant by the agreement being “an essential prerequisite” was that it was “a necessary step in the scheme” whereby the prospective profits were to be passed back (through Opendy) to the trustees. It was not being said that it was “necessary in any commercial sense”. Once this was properly understood, the Commissioners’ finding was fatal to the taxpayers’ claim:

“To have found that to agree to pay £77,250 for the benefit of an agreement which barely a week earlier had been assigned for £2,250 was a commercial purpose would have been simply perverse. After all, the directors of [the relevant company] had considered that, on March 30, 1962, £2,250 was a good price fully reflecting the value of the building agreement. The price Kilmorie paid was 34 times that good price.”

[490]Again, it is clear that Lord Wilberforce is referring to the lack of commerciality of the price in the context of justifying the Commissioners’ finding that the payment was made with a tax avoidance purpose.

[491]Lord Cross thought that the Special Commissioners could be criticised in so far as their language could suggest that the fact that one of the purposes for which a payment is made is not a trading purpose necessarily leads to the conclusion that the payment must be disallowed. He gave an example where a payment could be held to be for dual purposes but it was nevertheless allowable as a bona fide commercial transaction:

“Suppose that a retailer is in the habit of buying certain articles from a wholesaler for £10 each which is a fair commercial price, that his son-in-law sets up in business as a wholesaler dealing in similar articles and that thenceforth the retailer deserts the other wholesaler and buys the articles from his son-in-law for £10 each. One of the purposes for which the retailer is entering into the transactions with his son-in-law is to help him in business but nevertheless the cost would be properly allowable because the transactions though entered into in a sense for a dual purpose are bona fide commercial transactions.”

[492]However, although the language used could be open to misunderstanding, he thought the Commissioners were correct in their conclusion. Continuing with his example he put another case where the retailer bought articles from his son-in-law for £15 which he could have bought from the wholesaler for £10. He said that: “each expense would not have been allowable - at all events to the extent of the extra £5 - because the purchases were not genuine commercial transactions but purchases at a fancy price entered into to benefit the vendor”.

[493]Lord Cross noted that the benefit of the agreement for which Kilmorie paid over £77,000 had been sold for £2,250 only a few days previously and Mr Downes himself had said in evidence that that was a fair price. So “£77,000 was in truth a fancy price fixed by Downes and [the designers of the scheme] for the purposes of the scheme” – in the same way as £15 was a fancy price in his second example.

[494]He noted that the taxpayer stressed the fact that although Kilmorie had to pay £77,000 for the benefit of the agreement, it nevertheless derived a substantial profit from the transaction and also on the fact that the Special Commissioners found that the agreement “was an essential prerequisite to the carrying out by Kilmorie of the development of the estate.” He said:

“But these facts do not show that the price of £77,000 was a commercial price. It is, of course, true that Kilmorie could not develop the estate unless it acquired the benefit of the agreement from Opendy and that in order to acquire it had to pay £77,000. Further, it is true that the fact that a price paid is extravagant does not necessarily show that the purchase is not a genuine commercial transaction. A purchaser dealing at arm’s length with a vendor may say to himself ‘The price which he is asking is absurdly high but I cannot get him to take less and I believe that even at that price I can make a profit on the deal. So I will agree to pay what he is asking.’ But Kilmorie was not dealing at arm’s length with Opendy. It was controlled by Downes and it agreed to pay the £77,000 not because its directors other than Downes decided in the exercise of an independent judgment that it was worth Kilmorie’s while to agree to pay that price but because the scheme provided for that price being paid. For these reasons I would dismiss the appeal by Kilmorie.”

Scotts Atlantic

[495]Scotts Atlantic concerned a scheme designed to enable the taxpayers to provide their directors and employees with benefits, which were not taxable at that time, whilst obtaining a tax deduction for the cost of the scheme. It was designed to circumvent the provisions of schedule 24 to the Finance Act 2003 which are aimed at preventing exactly that situation.

[496]The Upper Tribunal decided that the transaction was not outside the scope of schedule 24 so that it applied to prevent there being any deduction which would otherwise be available. They went on, however, to consider the question of whether a deduction would be allowable under s 74 ICTA as, in effect, that was a pre-requisite to schedule 24 applying. On that issue, the taxpayer argued that the only purpose of the contributions was to provide benefits for employees through employee benefit trust arrangements. HMRC argued that there was another purpose, namely to secure a tax deduction. Not only was securing this tax deduction a purpose of the scheme (which was accepted) but it was also a purpose of the expenditure so that the contributions were not wholly and exclusively expended for the purposes of the trade.

[497]The Upper Tribunal set out the principles of the “wholly and exclusively” test adopting the principles set out in Vodafone including the need to distinguish between object and effect. They noted, at [55], that the mere fact that a choice is influenced or dictated by the tax consequences does not necessarily mean that the choice involves a duality of purpose as regards the expense. In their view: “the words of Millett LJ [as regards the distinction between object and effect] are just as relevant and applicable where there is a choice as where there is not: in each case, the question is whether the payment is made exclusively for the purposes of the trade, and that is a question of fact for the FTT”.

[498]At [56], the Upper Tribunal continued to note that the taxpayer had two arguments that the findings made by the tribunal in its decision in favour of HMRC was wrong. (1) The first argument was that the tribunal’s conclusion, that a main purpose of the relevant expense was to reduce tax liabilities, is inconsistent with the principle, as stated by Romer LJ in Bentleys, Stokes & Lowless v Beeson

[1952] 2 All ER 82

at 85, 33 TC 491 at 504 (and approved by the Court of Appeal in Interfish Ltd v Revenue and Customs Comrs

[2014] EWCA Civ 876

[2015] STC 55

) that:

“if, in truth, the sole object is business promotion, the expenditure is not disqualified because the nature of the activity necessarily involves some other result, or the attainment or furtherance of some other objective, since the latter result or objective is necessarily inherent in the act.”

The taxpayer noted that the tribunal had accepted the purpose of the expense was to benefit employees which was a business purpose. In the taxpayer’s view, the fact that the employees’ tax liability was reduced was merely an incidental effect and not a purpose. (2) The second argument, at [57], was that, even if there was an element of duality of purpose, the expenditure would have been incurred even without a non-trading tax avoidance motive and, relying on Kilmorie, in such circumstances the expenditure remains deductible.

[499]On the first issue, the Upper Tribunal considered precisely what the relevant finding of fact was that the tribunal had made. They noted, at [65] and [66], that a finding that there was an object of avoiding corporation tax would not be justified on the basis that the taxpayer had decided to incur expenditure for the purposes of its trade in a particular way. That would “confuse the object of the expenditure with the reasons for incurring it in the way in which it was in fact incurred”, as they had noted at [55], and, at [66]:

“A taxpayer is entitled to order its affairs in a way which incurs the least tax liability and the mere fact that a choice is influenced or dictated by the tax consequences does not necessarily mean that the choice involves a duality of purpose. It does not, therefore, necessarily follow that the adoption of the scheme … results in a duality of purpose (although it may do so as a matter of fact) unless this is one of those cases referred to by Lord Oliver in MacKinlay v Arthur Young … where the results (in the present case, the securing of deductions) are so inevitably and inextricably involved in particular activities (in the present case, the making of the contribution and the effecting of the scheme) that they cannot but be said to be a purpose of those activities.”

[500]However, the Upper Tribunal concluded that was not the basis for the tribunal’s finding. They noted, at [59], the distinction the tribunal drew between a case where a taxpayer ordinarily pays salary, in which case the fact that it expects to secure a deduction does not occasion any duality of purpose and, at [61], that where “the deliberate and all-pervading objective of achieving a corporation tax deduction makes it impossible to treat the corporation tax result sought for the contributions as the “ordinary, intended or realistically expected outcome” of making salary, bonus or equivalent payments. In their view, at [67], it was this last statement which essentially provided the basis for the tribunal’s findings. The words “ordinary, intended or realistically expected outcome” were adopted as the yardstick by reference to which a result can be ignored as a separate object.

[501]They were satisfied, at [70] to [73], that the tribunal had based its decision on a finding that they were entitled to make, that one of the purposes of the contributions (in contrast with the purpose of the method of expenditure) was to obtain a tax deduction which would not have been available if the contribution had been made by more conventional means, and that, in these particular circumstances, such purpose was not an incidental consequence of the expense. Whatever else, “the FTT did not conclude that because the tax benefit was a consequence of the contribution, it was a purpose of the expense.”

[502]They concluded by emphasising, at [73], that the important words in Romer LJ’s comments in Bentleys are “if, in truth, the sole object is business promotion”. In their view the principle is that “if such is the case, then an incidental effect does not as a matter of law disturb that conclusion”. In this case, the tribunal did not find that there was a sole object but only that the companies had another object. In doing so, it recognised that an incidental consequence of a tax deduction did not mean that such a deduction was necessarily an object. They continued, at [74], that the point made in Bentleys was that expenditure is not disqualified:

“because the nature of the activity necessarily involved some other result, in other words that the mere existence or knowledge of that result is not enough to give a dual purpose. But if the fact-finding tribunal concludes that its inquiry into the mind of the taxpayer revealed that the taxpayer actually had that other purpose as an object of the expenditure, then the fact that that result is a natural consequence of the expenditure will not cause that finding to be perverse.”

[503]It is clear, therefore, that the Upper Tribunal considered that the fact that a payment has tax consequences or that the way expenditure is incurred may result in a fiscal advantage does not necessarily result in the payment having a non-trading purpose. The question, as set out in Vodafone, must be assessed by reference to the particular object the taxpayer has in mind in making the relevant payment. If, on the subjective enquiry required, it is properly concluded that the taxpayer had in mind obtaining the tax benefit that results from the making of the payment, the fact that the obtaining of the tax benefit is a natural consequence of the payment does not affect that finding. In other words, there is no presumption that the natural consequence or effect of a payment is or is not the object of the payment.

[504]On the second argument the Upper Tribunal noted, at [75], that the taxpayer argued (similarly to BNP in this case) that Kilmorie shows that if the cost of something would have been £10 and its cost with tax benefit is also £10, then £10 remains deductible. This was said to follow both from the example in Lord Cross’s speech about a retailer and also from the division between the allowable and unallowable elements of an expense incurred with a dual purpose espoused by Lord Wilberforce in his agreement with Roskill LJ (

[1973] STC 330

[1973] 1 WLR 1180

) and by Lord Simon in his speech.

[505]They noted, however, at [76], that in that case none of their Lordships reached a conclusion that part of the payment should be allowed and part disallowed. The closest to that proposition were Lord Reid and Lord Cross:

“Further the House was considering the case in which it was possible to divide and segregate the expense into parts incurred for one purpose and parts for another, and Roskill LJ’s remarks in the Court of Appeal were also addressed to such a situation. That is quite different from saying that, if the whole of an expense had both a trading and a non-trading purpose, the existence of a trading purpose was sufficient to make the whole expense deductible. Such a result would seem to us to fly in the face of the statutory requirement that the purpose be ‘exclusively’ a trading one.”

[506]They continued, at [77], that they were “not deflected from this conclusion” by Lord Cross’s example of a retailer who buys goods for £10 from his son-in-law rather than for the same price from his usual supplier as set out above. They noted that in that case Lord Cross did not consider that the fact that the retailer purchased from the son-in-law to help him in his business prevented a deduction:

“But, whether the retailer buys for £10 from the previous wholesaler or his son-in-law he incurs the same expenditure, and, absent some particular factor, the object of that expenditure will almost certainly be found as a matter of fact to be to obtain the goods, even though the manner of the expenditure may be to benefit his son-in-law.”

[507]They then considered the other example given by Lord Cross where the retailer pays £15 to obtain the goods from his son-in-law when he could have paid £10 to his usual wholesaler. They said that in that case:

“it will generally be an obvious deduction from the circumstances that as a matter of fact he must have had an additional object in incurring the expenditure and not merely a different object in the manner it was incurred. But neither conclusion follows as a matter of law, for the actual evidence before the tribunal may dictate or permit a different factual conclusion.”

[508]In Drummond the Court of Appeal rejected HMRC’s argument that the whole of the price of £1.96 million paid by a taxpayer to acquire second-hand life assurance policies was not “consideration given wholly and exclusively” for the assets, within the meaning of the capital gains legislation, because the policies were purchased as part of a tax avoidance scheme to realise a loss on the policies. The Court of Appeal essentially agreed with Norris J in the High Court (Drummond v Revenue and Customs Commissioners

[2008] STC 2707

) who held (at [29] of his decision) that the taxpayer “wanted to acquire the policies precisely because by doing so he thought he would obtain a tax advantage on their surrender” but he nevertheless “still gave consideration wholly and exclusively to acquire them.” Discussion

[509]BNP submitted that in Kilmorie their Lordships (particularly Lords Reid, Cross and Wilberforce) did not consider the tax avoidance purpose to be determinative of the issue. Rather (with the exception of Lord Morris) they were focussed on the relationship between the amount expended and the value of what was acquired; the overpayment was the key to the issue. The corollary is that where, as here, the taxpayer pays a fair value to acquire a real asset, such as the right to a dividend, a tax deduction is available.

[510]BNP argued that the factual basis of the decision in Scotts Atlantic is distinguishable from this case and the reasoning used in fact supports BNP’s position. BNP noted that it was held in that case that the over-arching object of the payment was simply to get a deduction for the payment notwithstanding the enactment of schedule 24. That is not the same as a case where money is laid out to acquire an actual asset, as trading stock, for a price which reflects its fair value, for the purpose of onward sale. As the London branch paid a fair or market price for the dividend rights, it follows that the whole of that expenditure was incurred with the object of acquiring the dividend rights. This stands to reason given that, as a result of that expenditure, the London branch acquired an asset worth the price it paid. The ability to deduct the expense of acquiring trading stock is merely the natural consequence or effect of purchasing it at a fair or market value.

[511]BNP argued that, in their comments on Kilmorie, the Upper Tribunal in effect accepted that was a case where it was possible to divide and segregate the expenditure incurred on an asset into parts incurred for one purpose and parts for another and, by implication, that the division is to be made by reference to the value of the asset acquired. This is in contrast to a case, such as Scotts Atlantic, where nothing is acquired by the trader and there is no basis for any such division. The Upper Tribunal meant, therefore, that in Kilmorie the expenditure which reflected the fair value of the asset was incurred for the purposes of the trade; it was the remainder which was incurred for a non-trading purpose. In BNP’s view this is confirmed by the discussion at [77] of the example given by Lord Cross in Kilmorie of the case where the trader pays £10 to his son-in-law for goods worth £10 and the comment that “the object of that expenditure will almost certainly be found as a matter of fact to be to obtain the goods”.

[512]In our view, whilst in Kilmorie their Lordships refer to the lack of the commerciality of the price as a factor, that is quite clearly in the context of and leading to the conclusion that the over payment demonstrated that the taxpayer’s object was to facilitate a tax avoidance scheme. It was not simply the fact that the price was inflated which lead to their conclusion that the expenditure was not allowable. It was the fact that the inflated price showed that there was a non-commercial purpose which, in the context of the Commissioners’ findings of fact, lead to that conclusion that there was a tax avoidance purpose.

[513]As for the examples given by Lord Cross, read in context, Lord Cross was merely drawing attention to the fact that the particular object of the payment must be properly assessed in each case. It is not every ancillary non-trading purpose which will prevent a payment qualifying as wholly and exclusively incurred for trading purposes. In other words, buying an asset from a relative for its usual price will not convert the object of the payment from being for commercial trading purposes to non-trading purposes. On the other hand, paying a relative an overvalue for the asset, as he continued to note in his second example, may (as in Kilmorie itself) lead to the conclusion that the object of the payment is not in fact wholly commercial. In the context of examining whether the Commissioners had made correct conclusions that the object was in part for tax avoidance purposes, Lord Cross clearly considered that the over inflated price paid in that case was, in effect, evidence of that object. So he concluded that Kilmorie paid that price not because it was worth its while “but because the scheme provided for that price being paid”.

[514]We do not see this decision, therefore, as establishing a principle, as BNP argued, that if a fair price is paid for an asset of a kind used in the taxpayer’s trade, that amount will necessarily qualify as deductible expenditure. In each case, the question is why, or with what object or purpose, the amount was laid out or expended. Whilst expenditure incurred in such circumstances ordinarily may well be incurred wholly and exclusively for the purposes of the trade, the decision does not establish that there is a presumption to that effect. There is still scope for finding that the price was not paid for trading purposes if that was not in fact the case. The focus on the over inflated price paid by the taxpayer in Kilmorie was referred to as the main factor demonstrating that a purpose or object in paying the amount in that case was to facilitate the tax avoidance scheme.

[515]HMRC argued that there is no support for BNP’s position in the Upper Tribunal’s comments on Kilmorie in the Scotts Atlantic case. We agree. The Upper Tribunal recognised that there is a difference between a case where a payment can be separated into parts each of which has a separate purpose and that where the whole payment has two purposes (see [76]). They saw Lord Cross’s first example as one in which “the object of that expenditure will almost certainly be found as a matter of fact to be to obtain the goods” and the second as one in which it “will generally be an obvious deduction from the circumstances that as a matter of fact he must have had an additional object in incurring the expenditure” but “neither conclusion follows as a matter of law, for the actual evidence before the tribunal may dictate or permit a different factual conclusion” (see [506] and [507] above).

[516]All the Upper Tribunal drew from Lord Cross’ examples, therefore, was that it is purely a question of fact as to what is the object of the expenditure. Lord Cross’ examples were just that; examples of what might be expected to be the case. They were not put forward as establishing legal principles. On that basis, we cannot see that the Upper Tribunal was, as BNP asserted, endorsing a proposition that an amount spent on an asset must necessarily be deductible if the amount reflects its value. Rather once it has been found as a fact that the expenditure has two purposes, one being a business purpose and one not, then the expense is not deductible.

[517]Moreover, the Upper Tribunal’s conclusions on the taxpayer’s first argument in Scotts Atlantic support this conclusion. They considered that the passage cited in Bentleys meant that it is not to be presumed that, if an effect necessarily follows from a payment, then the effect is an object of that payment. However, if in fact, on the subjective enquiry required, it is properly found to be what the taxpayer had in mind as the object of the payment, it is not precluded from being such because it is a necessary consequence of the payment. Hence, the fact that a contribution payment of the type involved in that case would usually result in a tax deduction neither leads to a conclusion that obtaining the deduction is a mere incidental effect of the payment nor that it is the object of the payment. A further enquiry is required. As in that case, therefore, the fact that it would usually follow as a necessary consequence that a trader who buys trading stock for its fair value and sells it for a profit obtains a tax deduction does not of itself lead to the conclusion that the payment is deductible. If, in truth, the appellant’s purpose in incurring the expenditure was to obtain a tax advantage, the tribunal is not precluded from finding that was the taxpayer’s object.

[518]BNP also relied on Drummond but we agree with HMRC that the case is not in point. The question for capital gains purposes is whether the monies have been given (wholly and exclusively) as consideration for the acquisition of the asset. Hence, whilst the purpose of the taxpayer in Drummond in acquiring the insurance policies may well have been because he wanted to implement a tax avoidance scheme, that was simply not relevant. It seems to us that is not the same question as whether monies have been laid out wholly or exclusively for the purposes of the trade which, as set out in the case law, is a subjective question of looking at the taxpayer’s intentions. The question what were the monies spent on is not the same as why were the monies spent.

[519]BNP also argued that expenditure which is incurred wholly and exclusively for the purposes of a trade does not cease to be deductible simply because, in the particular circumstances, the receipt “generated” by the expenditure is not brought into the trader’s tax computation on the basis of Hughes v Bank of New Zealand

[1938] AC 36

(‘Hughes’). Whilst that may well be the case, as HMRC noted, that decision sheds no light on when an expense is to be regarded as wholly and exclusively incurred for the purposes of the trade in the first place. At 378, Lord Tankerton merely acknowledged that expenditure in the course of the trade which is unremunerative is none the less a proper deduction, if wholly and exclusively made for the purposes of the trade. It does not require the presence of a receipt on the credit side to justify the deduction of an expense.

[520]As BNP repeatedly emphasised, the London branch paid a price for the right to the dividends which reflected a fair value for the right (or at least the price was not an overpayment looking at what a third party may be expected to pay), the right is analogous to assets the London branch typically deals in and it was sold for a profit. In one sense, it can be said that monies spent on an asset typically bought and sold in the trade for its fair value are laid out for acquiring the asset which, as a factual matter, is thereby obtained. Whilst that may often lead to the conclusion that the asset was acquired for trading purposes, that is not the end of the enquiry and any wider purpose may be relevant. As the Upper Tribunal said in Scotts Atlantic, it does not necessarily follow that because an effect follows as a natural consequence of the incurring of expenditure that that result is not a purpose of the expenditure. So the fact that it follows in the usual course that spending monies on an asset which is sold in the course of the trade gives rise to a tax deduction does not preclude that effect being the taxpayer’s object if an enquiry into the taxpayer’s mind reveals it to be such.

[521]In this case, as we have set out in detail above, we consider it clear on an objective assessment that the taxpayer’s aim in undertaking the transaction was to obtain the s 730 benefit. Those factors also demonstrate in our view that the London branch’s objective in incurring the expenditure on the right to the dividend was to access the s 730 benefit. This was not a case of the London branch paying £149.1 million with the aim of acquiring an asset for sale in its trade to generate a commercial profit but of spending that amount to obtain a tax benefit with the acquisition and sale of the asset (as designed to achieve a profit) as the necessary steps required to achieve that.

[522]That this was the aim of those involved in the project in a subjective sense is very clear from the internal correspondence, the papers prepared for the approvals process and the minutes of the TCC. We note, in particular, the email from Mr Neil Robinson dated 26 October 2005 extolling the tax benefits of the scheme, the email in a similar vein from Mr Scholes dated 11 November 2005 and that of Mr Robinson dated 8 December 2005 noting that “in an uncertain environment where disclosure can lead to a change in law it is critical that we try and source principal deduct transactions (such as Sumatra) when looking to optimize the tax position of the Bank and our clients in the UK” (see [54], [71] and [97]).

[523]Whilst there were commercial effects of the transaction, these were introduced and designed with the objective of enhancing the prospect of achieving the s 730 benefit. Whilst the papers prepared for internal approval purposes refer in headline terms to the benefit of the so called “cheap” funding, the focus of the papers is on the tax effect of the transaction. The minutes of the meeting of the TCC reveal no consideration of the commercial merits of the transaction other than the obtaining of the tax benefit. The correspondence and the papers acknowledge that, in fact, the benefit for the BNP group was obtained immediately such that it would be in its interests to terminate the arrangements with ALIL immediately (see [106]). The OF team were remunerated by reference to the potential tax benefit. Mr Demon and Mr Peters both accepted that the transaction would not have happened but for the potential obtaining of the tax benefit and that it was least a main purpose of the transaction. Accordingly, we conclude that the expenditure incurred by the London branch in acquiring the dividend rights from BNP Lux was not incurred wholly and exclusively for the purposes of the trade of the London branch but (at least in part if not solely) for the purpose of obtaining the s 730 benefit for the BNP Group. Closure notice issue

[524]The next issue is whether HMRC are entitled to advance their argument on s 730(3). The amendment which is the subject of the appeal was made by HMRC on the closure of their enquiries into the taxpayer’s position in the period in question in a notice dated 22 October 2010. The dispute is whether this issue falls within the scope of the conclusions and amendments made on issue of the closure notice and thereby within the scope of the appeal proceedings. The parties differ on the correct approach to this issue on the basis of the decision of the Supreme Court in Tower MCashback LLP 1 and Another v Revenue and Customs Commissioners

[2011] 2 AC 457

and

[2011] STC 1143

and more recently by the Court of Appeal in the Fidex case. Statutory provisions

[525]HMRC may enquire into a company’s corporation tax return under para 24 of schedule 18 to the Finance Act 1998. The provisions in schedule 18 relating to the completion of an enquiry state the following:(1) “An enquiry is completed when [an officer of Revenue and Customs] by notice (a “closure notice”) [informs] the company they have completed their enquiry and state their conclusions. The notice takes effect when it is issued” (para 32).

(2) The closure notice must: “(a) state that, in the officer’s opinion, no amendment is required to of the return that was the subject of the enquiry, or (b) make the amendments of that return that are required - (i) to give effect to the conclusions stated in the notice …” (para 34(2))

(3) “An appeal may be brought against an amendment of the company’s return under sub-paragraph (2).” (para 34(3))

[526]The tribunal has power, on an appeal made to it, to reduce the amount of an assessment or otherwise the assessment “shall stand good” under s 50 of the Taxes Management Act 1970. Closure notice and related correspondence

[527]In their closure notice issued on 22 October 2010 HMRC said the following:

“I have completed my enquiry into the company’s Tax return for the period 1 January 2005 to 31 December 2005 and my conclusion is as follows: - [The London branch] is not entitled to a deduction for the amount payable by it to acquire from [BNP Lux] the right to dividends from [HIL]. My calculations are as follows Loss for the period based on return as amended £91,091,000 Acquisition costs of the strip (as D31 in comps) £149,106,000 Revised profit £58,015,000 Tax due at 30% £17,404,500 This notice amends the return to give effect to my conclusions.”

[528]HMRC sent a letter also dated 22 October 2010 (the “accompanying letter”) setting out an explanation of their position. This stated the following at the start of the letter:

“Our contention is that the purchase and subsequent sale of the dividend rights by [the London branch] should both be disregarded for taxation purposes. The return for 2005 as made excludes from taxation the proceeds of the sale of the rights so the adjustment required is to disallow the acquisition costs. We have alternative grounds for this adjustment, which we have set out below under the headings Ramsay and Lupton.”

[529]The Lupton issue referred to was the argument considered above that the purchase and sale of the right to the dividends was not a trading transaction. The Ramsay issue was that “when a realistic view is taken of the facts, the transaction in question does not fall within s 730 purposively construed”.

[530]To explain the “Ramsay” issue further, HMRC quoted quite extensively from the cases on the approach to purposive construction. They then set out the provisions of s 730 and what they considered to be its purpose and specifically the purpose of s 730(3), being to avoid double taxation. Having considered the commerciality (or rather in their view, the lack of commerciality) of the transaction HMRC concluded that: “the interposition of [the London branch] into the pre-planned transactions was solely for tax purposes and the evidence supports that [the London branch] hoped to obtain a trading deduction for the cost of purchasing the rights to dividends from BNP Lux whilst claiming that s 730(3) applied to the sale of rights to dividends so that it was not taxed on the sum received from ALIL … There are two parts to our argument based on our construction of s 730(3). The first is a consideration of what is a realistic view of the facts bearing in mind that that it is permissible to consider a pre-planned series of transactions as a composite and secondly whether in considering the purpose of the legislation here there is scope for the role [of the London branch] to be ignored on the authority of Lord Hoffman in Macniven at para 48… We say that taking a realistic view of the facts section 730(3) is not engaged because there is no prospect of double taxation as a result of the pre-planned composite transaction”.

[531]HMRC then set out in the accompanying letter the amendments required to the appellant’s corporation tax return (by adding the acquisition cost of the dividends back into the computation as set out in the closure notice). They concluded by noting that they were not pursuing two previous lines of argument including the argument now raised as regards s 730(3):

“Other Arguments In our letter of 3 April 2009 we advanced two other arguments, namely that the application of the legislation at Section 730 ICTA 1988 does not permit [the London branch] to treat the sale proceeds from the dividend rights as non-taxable receipts and that if the £150,000,000 receipt is not taxable as a trading receipt by virtue of S 730(3) ICTA 1988 then it would be chargeable to corporation tax as a chargeable gain. We have decided not to pursue these contentions.”

Case law

[532]In Tower MCashback, a limited liability partnership, claimed capital allowances under s 45 of the Capital Allowances Act 2001 in relation to expenditure it claimed it had incurred on a software licence agreement. HMRC opened an enquiry in which they focused on a particular provision in s 45, being s 45(4), which withholds first year allowances for expenditure on software rights “if the person incurring it does so with a view to granting another person a right to use or otherwise deal with any of the software in question”. In their closure notice, HMRC said “as previously indicated my conclusion is: the claim for relief under section 45 is excessive”. The closure notice was sent with a covering letter which stated:

“I am satisfied that the Tower MCashback scheme fails on the section 45(4) point alone.”

[533]HMRC later wanted to rely on a new argument, namely, that the taxpayer had not incurred the expenditure in buying the software licence within the meaning of s 45 because over 75% of the funds needed for the purchase had been borrowed against security provided by the seller on uncommercial terms. The Special Commissioner ruled in favour of HMRC that he had jurisdiction to consider the new argument. Mr Justice Henderson reversed this in the High Court (HMRC v Tower MCashback LLP 1 and Another

[2008] EWHC 2387

[2008] STC 3366

) but the Court of Appeal (HMRC v Tower MCashback LLP 1 and Another

[2010] EWCA Civ 32

[2010] STC 809

) and the Supreme Court decided in favour of HMRC.

[534]Mr Justice Henderson made a number of comments which were referred to by the Supreme Court:(1) At [113] he noted that there was no express requirement for the officer to set out his reasons for his conclusions and that what mattered “is the conclusion which the officer has reached upon completion of his investigation of the matters in dispute, not the process of reasoning by which he has reached those conclusions.”

(2) He said, at [115], that there is a principle of tax law “to the general effect that there is a public interest in taxpayers paying the correct amount of tax……[which] still has at least some residual vitality in the context of section 50” such that the Commissioners must:

“be free in principle to entertain legal argument which played no part in reaching the conclusions set out in the closure notice. Subject always to requirements of fairness and proper case management, such fresh arguments may be advanced by either side or may be introduced by the Commissioners on their own initiative.” (3) He then said, at [116], that this did not mean that an appeal against a closure notice “opens the door to general roving enquiry into the relevant tax return”. Rather: “The scope and subject matter of the appeal will be defined by the conclusion stated in the closure notice and by the amendments (if any) made to the return. The legislation does not say this in so many words, but it follows from the fact that the taxpayer’s right of appeal under section 31(1)(b) is confined to an appeal against any conclusion stated or amendments made by a closure notice. That is the only appeal which the Commissioners had jurisdiction to entertain.”

[535]At [128], he noted that “the result may from the Revenue’s point of view be characterised as conferring a windfall benefit on the taxpayer” but that another way of looking at the limitation on the scope of the appeal is as “part of the protection given by Parliament to taxpayers under the self-assessment system. There is always a balance to be struck between the interest of individual taxpayers on the one hand and the interest of the State and the general body of taxpayers on the other hand. Parliament has decreed how the balance is to be struck…”.

[536]In the Court of Appeal, the majority of Moses LJ and Scott Baker LJ (with Arden LJ dissenting) largely appeared to agree with the reasoning of Henderson J but reached a different conclusion in applying those principles. Moses LJ also noted the public interest in the correct amount of tax being paid and, at [29], that the self-assessment regime contains a system of checks and balances which, at [31], it is not to be supposed that Parliament intended to be overridden by the retention of a system of “thoroughly uninformative notices of assessment and notices of appeal”. He noted, at [31] and [32], that HMRC accepted some restriction being that it is implicit in the statutory scheme that an appeal “is confined to the subject matter of the conclusions and any amendments stated in the notice” (as was held by Dr John Avery Jones CBE in D’Arcy v Revenue and Customs Comrs [2006] STC (SCD) 543) but, at [33], “it all depends what one means by the ‘subject-matter’.”

[537]Moses LJ referred, at [34], to Henderson J’s comments at [113] and [116] and concluded, at [35], that he was driven (by the relevant provisions in the context of the restrictions imposed on HMRC’s power to amend a self-assessment) to the same view as Henderson J:

“The subject matter of this appeal is defined by the subject matter of the enquiry and the subject matter of the conclusions which close that enquiry. But that statement of principles serves only to give rise to further questions and problems. As this appeal demonstrates, there is likely to be controversy as to how one draws the boundaries of the subject matter of the conclusions stated in the closure notice. Are reasons for the conclusion to be distinguished from the conclusion stated, and if so, how?”

[538]At [37], he warned against too rigid an approach noting that, as Parliament had not chosen to identify some legal principle on this issue, it would be wrong for the court to attempt to do so and any such statement of principle “is likely to condemn both taxpayer and the Revenue to too rigid a straitjacket” and may “prevent a taxpayer from advancing a legitimate factual or legal argument which had hitherto escaped him or deprive, on the other hand, the public of the tax to which it is entitled.”

[539]At [38], he said that “with those nebulous observations” he would leave it to the Commissioners (now the tribunal) to identify the subject matter of the enquiry and thus the subject matter of the conclusions in which exercise the tribunal will have to “balance the need to preserve the statutory protection for the taxpayer afforded by notification that the inspector has completed his enquiries and the need to ensure that the public are not wrongly deprived of contributions to the fisc.”

[540]He continued, at [41], to state that it is to the tribunal that the statute looks to identify what s 28ZA describes as the subject matter of the enquiry:

“The closure notice completes that enquiry and states the inspector’s conclusions as to the subject matter of that enquiry. The appeal against the conclusions is confined to the subject matter of the enquiry and of the conclusions. But I emphasise that the jurisdiction of the Special Commissioners is not limited to the issue whether the reason for the conclusion is correct. Accordingly, any evidence or any legal argument relevant to the subject matter may be entertained by the Special Commissioner subject only to his obligation to ensure a fair hearing.”

[541]At [42], he expanded on this as follows:

“Protection of the public requires, at the least, that other issues arising from the subject matter of the enquiry ought to be considered, if necessary, by the fact-finding tribunal. In D’Arcy [2006] STC (SCD) 543 at para 11, the Special Commissioner ruled that the scope of an appeal against a conclusion or amendment made by a closure notice will depend on the facts. The conclusion in that case was, as described by Dr Avery Jones, very specific and relied upon the Ramsay principle. But the Special Commissioner permitted other issues arising from the facts to be advanced since the tribunal must form its own view on the law without being restricted to what the Revenue stated in their conclusion or the taxpayer states in the notice of appeal (see para 13). I see no reason for confining that view merely to legal issues. Provided a party can be protected from ambush, the only limitation on issues which might be entertained by the Special Commissioner is that those issues must arise out of the subject matter of the enquiry and consequently its conclusion, and be subject to the case management powers to which I have referred.”

[542]Lord Justice Moses concluded, at [51], that the closure notice did not of itself allow so restricted a view of the subject matter of the appeal as had been decided by Henderson J. Whilst “it did refer to previous correspondence which clearly focused on section 45(4), the closure notice itself was in plain terms a refusal of the claim for relief under section 45”.

[543]In the Supreme Court, Lord Walker, at [15], approved Henderson J’s comments at [113], [115] and [116] of his judgment noting that he had reached his conclusion “despite having correctly made” those observations. He then referred, at [16], to the comments of Moses LJ, at [32] and [41] of his decision, concluding, at [17], that there was “little if any difference” between the majority of the Court of Appeal and Henderson J as to the principles to be applied; the difference was as to the application of those principles. He preferred the approach of Moses LJ.

[544]Lord Walker cautioned, at [18], against the decision being taken as encouragement to draft every closure notice in wide and uninformative terms although, “if, as in the present case, the facts are complicated and have not been fully investigated, and if their analysis is controversial, the public interest may require the notice to be expressed in more general terms”. As both Henderson J and the Court of Appeal observed, unfairness to the taxpayer can be avoided by proper case management during the course of the appeal. He noted that similarly, Dr Avery Jones observed in D’Arcy v HMRC [2006] STC (SCD) 543, para 1:

“‘It seems to me inherent in the appeal system that the tribunal must form its own view on the law without being restricted to what the Revenue state in their conclusion or the taxpayer states in the notice of appeal. It follows that either party can (and in practice frequently does) change their legal arguments. Clearly any such change of argument must not ambush the taxpayer and it is the job of the Commissioners hearing the appeal to prevent this by case management.’”

[545]Lord Hope said, at [83], that, as the right of appeal under the relevant provision is against the conclusion stated in or amendment made by a closure notice, “it is desirable that the statement by the officer of his conclusions should be as informative as possible”. He noted that the closure notice was in very bald terms and whilst “the statute does not spell out exactly what it means by the words ‘his conclusions’ … taxpayers are entitled to expect a closure notice to be more informative”.

[546]He continued, at [84], that such notices “are seldom, if ever, sent without some previous indication during the enquiry of the points that have attracted the officer’s attention. They must be read in their context.” In that case, as the officer drew attention to his previous indications and sent a covering letter which cast further light on the approach, he did not think that it was unfair to the taxpayers to hold that the issue as to their entitlement to allowances should be examined as widely as may be necessary to determine whether they are indeed entitled to what they have claimed. Furthermore:

“while the scope and subject matter of the appeal will be determined by the conclusions and the amendments made to the return, s 50 of TMA does not tie the hands of the commissioners (now the Tax Chamber) to the precise wording of the closure notice when hearing the appeal.”

[547]In Fidex, the issue was whether under a complex scheme, the appellant had generated an allowable loss of nearly €84 million where Fidex sought to obtain the benefit of a specific rule in the loan relationship provisions (para 19A in the relevant provisions). In their closure notice, HMRC referred only to their argument on that provision but later sought to argue that the debit was not allowable under a different provision in the loan relationship rules. It was held that the tribunal was right to conclude that the subject matter of the enquiry/the closure notice and of the review related to the admissibility of the debit claimed and that the new argument could be raised as an additional ground upholding that conclusion.

[548]In setting out a comprehensive review of Tower MCashback, Kitchin LJ noted, at [43], that the appellant drew attention to the use by Moses LJ in his judgment of the phrase “the subject matter of the enquiry” but, at [44], that he did not think Moses LJ had meant to expand the permissible scope of an appeal in his use of this phrase beyond that contemplated by Henderson J nor did he understand the Supreme Court to have sanctioned any such expansion. In his view, Moses LJ was, “doing no more than explaining that the closure notice must be considered in context and in light of the enquiry that preceded it”.

[549]He concluded, at [45], that the principles to be applied are those set out by Henderson J as approved by and elaborated upon by the Supreme Court which, so far as material to that appeal, may be summarised in the following propositions:

“i) The scope and subject matter of an appeal are defined by the conclusions stated in the closure notice and by the amendments required to give effect to those conclusions. ii) What matters are the conclusions set out in the closure notice, not the process of reasoning by which HMRC reached those conclusions. iii) The closure notice must be read in context in order properly to understand its meaning. iv) Subject always to the requirements of fairness and proper case management, HMRC can advance new arguments before the FTT to support the conclusions set out in the closure notice.”

[550]He noted, at [51], that in his view the Upper Tribunal had been right not to take too rigid an approach as though this was a question of statutory construction. He agreed with them that “it is not appropriate to construe a closure notice as if it is a statute or as though its conclusions, grounds and amendments are necessarily contained in watertight compartments, labelled accordingly”. He also thought they had rightly emphasised that “while there must be respect for the principle that the appeal does not provide an opportunity for a new roving enquiry into a company’s tax return”, the tribunal is not deprived of jurisdiction “where it reasonably concludes that a new issue raised on an appeal represents an alternative or an additional ground for supporting a conclusion in the closure notice”. He said, at [64], that just as Lord Hope observed in Tower MCashback, it was not unfair to Fidex to hold that the issue as to its entitlement to the debit should be examined as widely as might be necessary to determine whether it was indeed entitled to what it had claimed. Discussion on closure notice issue

[551]It is not disputed that it is clear from the principles set out in Tower MCashback and Fidex that the scope of the appeal in a case such as this is to be determined essentially by the subject matter of the conclusions and amendments set out in the closure notice. As Henderson J said, in the passages from his judgment expressly approved by Lord Walker in the Supreme Court, the tribunal cannot stray beyond that. The courts have emphasised that the tribunal is not merely acting as an arbiter between the parties but in the public interest in determining the correct amount of tax due. Inevitably, the exercise involves balancing the protection for the individual and the public interest in the payment of the right amount of tax. So, whilst the legislature has, as Henderson J noted, provided for the drawing of a line by reference to the subject matter of the conclusions, precisely where to draw that line, taking into account this balancing exercise, is left to the tribunal to determine on the facts of the particular case.

[552]We note that Lord Walker, like Moses LJ, approved the comments of Dr Avery Jones in the D’Arcy case that it is “inherent in the appeal system that the tribunal must form its own view on the law without being restricted to what the Revenue state in their conclusion or the taxpayer states in the notice of appeal” and that it followed that “either party can (and in practice frequently does) change their legal arguments” although “clearly any such change of argument must not ambush the taxpayer and it is the job of the commissioners hearing the appeal to prevent this by case management”.

[553]As Lord Hope said, closure notices “must be read in their context” and section 50 does “not tie” our hands “to the precise wording of the closure notice” when hearing the appeal. As the Upper Tribunal said, as approved by Kitchin LJ in the Court of Appeal in Fidex, it is not appropriate “to carry out this exercise as though it were a matter of statutory construction or as though its conclusions, grounds and amendments are necessarily contained in watertight compartments, labelled accordingly”.

[554]BNP argued that the statement by HMRC in the closure notice is the entirety of the subject matter of the conclusion, namely, that the price paid by the London branch for the dividend rights was not deductible. The amendment made by the exclusion of the purchase price from the tax computation as a tax deductible amount is entirely consistent with this conclusion. Moreover, HMRC expressly said in the accompanying letter that they were not pursuing the s 730(3) argument.

[555]BNP referred in particular to the passages from the judgment of Henderson J approved by Lord Walker in Tower MCashback and the summary set out by Kitchin LJ in Fidex. BNP asserted that, on that approach, whilst the assessment has to be made in context, it is the conclusion as set out in the closure notice which is paramount. In BNP’s view, Lord Justice Kitchen’s fourth proposition makes it clear that the question of jurisdiction and of fairness or case management are two entirely separate questions. It is only if the tribunal has jurisdiction to hear a particular argument, on the basis that it is within the subject matter of the conclusion, that any question of fairness or prejudice can arise. Questions of fairness do not arise here as that prior jurisdictional test is not satisfied.

[556]HMRC argued that, looking at the issue in the broader context of the correspondence, the subject matter of this appeal is the tax efficacy of the arrangements put in place by the BNP group. Section 730(3) has always been in issue; initially this was by reference to Ramsay as set out in the accompanying letter and review correspondence. There is no unfairness in requiring BNP to satisfy the tribunal on the s 730 score given the public interest in ensuring the correct payment of tax. The fact that this would involve altering the computations made in the closure notice is immaterial. These were produced on the basis of HMRC’s primary case that the transaction was not a trading transaction.

[557]HMRC referred to the approach set out by Moses LJ in the Court of Appeal in Tower MCashback. They considered that Lord Walker approved not just the conclusion of Moses LJ but also his reasoning. In their view, his judgment gives greater emphasis to the context, being in this case the relevant chain of correspondence leading up to the closure notice and in any review, and that this is not a question of arid construction. Moreover, it is clear that the public interest in the right amount of tax being paid is relevant to the question of the tribunal’s jurisdiction as regards the very scope of the appeal as well as to case management issues of fairness.

[558]They noted that the argument HMRC wish to raise does not raise any new issue of fact but is purely a question of law. Lord Walker suggested in Tower MCashback that, as is consistent with the tribunal’s duty to determine the correct amount of tax, the tribunal cannot be confined to whatever the parties have said about the correct legal interpretation of a particular provision. In determining the subject matter of this appeal, it is not possible to separate the sale of the right to the dividends from the purchase of those rights. The appellant has to show that this was a trading transaction by reference to both limbs of the transaction. The appellant’s scheme hinges, assuming this was a trading transaction, on the proceeds of the sale not being taxable under s 730(3). That is a pure question of the construction of s 730 as a question of law. It arises out of exactly the same facts which justify, on the appellant’s case, treating the expenditure on the acquisition of the dividend rights as trading expenditure. From the correspondence, it is clear that the grounds of dispute put by HMRC related to both limbs of the transaction.

[559]We note that the parties seem to suggest that there is a difference in approach between Henderson J and Moses LJ and as to whether the Supreme Court was fully endorsing the comments of Moses LJ. As raised in Fidex, Moses LJ referred several times to the need to have regard to the subject matter of the enquiry. At [41] of his judgment, he said that the tribunal must have regard to “the subject matter of the enquiry” as “the closure notice completes that enquiry and states the inspector’s conclusions as to the subject matter of that enquiry”. The only limitation, as he continued at [42], on what might be entertained “is that those issues must arise out of the subject matter of the enquiry and consequently its conclusion”.

[560]As Kitchin LJ said in Fidex, it does not seem, however, that Moses LJ intended to broaden the scope of the exercise beyond that described by Henderson J nor did he consider that the Supreme Court sanctioned any such extension. We note that Moses LJ was clearly of the view that the context of the subject matter of the enquiry may be highly relevant but, in each case, he focussed on the fact that the closure notice completes that enquiry. We take this to mean that the conclusions in the closure notice are to be informed by the subject matter of the previous enquiry, as the document bringing the enquiry to an end. This accords with the view of Lord Justice Kitchin that Moses LJ was simply doing no more than explaining that the closure notice must be considered in context and in light of the enquiry that preceded it. That Henderson J regarded context as relevant is inherent in his decision, which was essentially based on what was said in the surrounding correspondence, as limiting the scope to s 45(4).

[561]In broad terms, as is clear from the amendment made by the closure notice and the related correspondence, the matter in dispute was whether the transaction had created an allowable trading loss of over £91 million which could then be surrendered by way of group relief. The amendment made by HMRC wiped out the claimed loss and produced a profit. This was achieved by HMRC adding back into the computation the amount of just over £149 million spent by the London branch on acquiring the dividend rights. The issue is essentially whether the statement in the closure notice, that the amount spent on the dividend right was not deductible, represents the entirety of HMRC’s conclusion or is merely an argument or grounds for some broader conclusion.

[562]Having regard to the context in which the closure notice was issued and, in particular, the statements in the accompanying letter, we agree with HMRC that it is clear that the factual scope of the matter in dispute when the closure notice was issued was not confined to the purchase side of the transaction only. Indeed, as HMRC argued, by its very nature the scope of the subject matter cannot be so confined.

[563]HMRC concluded in the accompanying letter that the purchase and sale of the dividend rights should both be disregarded for taxation purposes. As the taxpayer’s computation did not take into account the sales proceeds, the adjustment required was to add back in the purchase price. The rest of the letter sets out the grounds for that conclusion being(a) that the purchase and sale was not a trading transaction (on the authority of Lupton) and

(b) alternatively that, as a matter of purposive statutory construction, on a realistic view of the facts, the purchase and sale of the dividend rights by the London branch was to be disregarded (the Ramsay argument). Under the Ramsay argument their view was that, on a purposive construction of s 730, the terms of s 730 (3) were never engaged as there was no question of double taxation (as set out in further detail above). Whilst we find HMRC’s Ramsay argument somewhat difficult to follow, it is clear that they were basing it, whatever the merits of the argument, on their interpretation of a purposive approach to the construction of s 730, including s 730(3), albeit that their view was that the provision was simply not engaged.

[564]It is clear from this correspondence that the focus of HMRC’s enquiry and of the conclusion they then reached was on whether the transaction, being the purchase and sale of the right to the dividends by the London branch, resulted in an allowable trading loss of over £91 million as the London branch claimed. At the point when the closure notice was issued, they had two reasons for asserting that it was not, both of which looked at the taxation of the transaction in its entirety.

[565]On that basis, we do not regard the conclusions or amendment set out in the closure notice, as assessed according to the context of the enquiry and related correspondence, to be confined to the narrow point that the London branch was not entitled to a tax deduction for the price paid for the dividend rights. Looked at in context, HMRC concluded the enquiry by taxing the transaction as though it were not a trading one or as though it is to be disregarded for fiscal purposes.

[566]The result, in each case, is that there is no trading loss. The mechanics of denying the loss were to add back into the computation the price paid by the London branch for the acquisition of the dividend rights. In our view, the argument that s 730(3) does not operate, as BNP asserted, to exclude the proceeds of the sale of the transaction from the corporation tax computation, is simply another argument which supports the conclusion that the transaction has not created an allowable loss. The fact that a different mechanic would be needed to achieve that result, if the argument were to succeed, by adding the proceeds of sale into the computation is immaterial given that the overall result would be the same.

[567]Given that, in our view, this argument can be raised within the scope of the conclusions embodied by the closure notice as interpreted in its context, the only remaining considerations are case management ones such as avoiding any ambush. However, there is no such concern in this case. The appellant has had sufficient time to consider this argument and indeed it does not raise any argument to the contrary. We also note that we do not consider that, as a matter of principle, there can be any difference between the raising of a wholly new argument and the reviving of an argument that had been raised but dropped. In each case, having decided that it is within the tribunal’s jurisdiction to hear the argument, the only considerations preventing it from being heard would be case management ones. Section 730 issue

[568]We turn finally to the substantive issue on the application of s 730(3). Essentially, the issue is whether s 730(3) applies to prevent the proceeds from the sale of the right to the dividends from being taxed in the hands of the London branch.

[569]Section 730 had its roots in s 24 of the Finance Act 1938 which, as was common ground, was enacted in response to the decision of the Court of Appeal in IRC v Paget 21 TC 677. In that case, it was held that the proceeds of sale of interest coupons on bonds held by Miss Paget as investments were not taxable as income. The provisions were anti-avoidance measures designed to prevent investors alienating rights to interest on securities (whilst retaining the securities themselves), normally for a capital sum, thereby avoiding the charge to income tax, which would be imposed on the receipt of the interest itself.

[570]That this was the intention behind s 730 is acknowledged in the explanatory notes to the draft legislation which, when subsequently enacted (as para 2 of Schedule 7 to the Finance (No 2) Act 2005), made a number of changes to s 730 to produce the version in force when the transaction took place (the “applicable version”). The notes state:

“Section 730 ICTA 1988 prevents avoidance of tax where a person sells or transfers the right to income from a security without selling or transferring the security itself. It reverses a court decision in the 1930s that the proceeds of sale of such income were not taxable.”

Provisions of s 730

[571]We have set out below the full text of the applicable version and highlighted the main differences with the version which was in force immediately before that (the “preceding version”), by including the previous wording in the applicable version in italics. “730 Transfers of rights to receive distributions in respect of shares(1) Where in any chargeable period the owner of any shares [securities] (“the owner”) sells or transfers the right to receive any distribution [interest] payable (whether before or after the sale or transfer) in respect of the shares [securities] without selling or transferring the shares [securities], then, for all the purposes of the Tax Acts, that distribution [interest], whether it would or would not be chargeable to tax apart from the provisions of this section – (a) shall be treated as [shall be deemed to be] the income of the owner or, in a case where the owner is not the beneficial owner of the shares [securities] and some other person (“a beneficiary”) is beneficially entitled to the income arising from the shares [securities], the income of the beneficiary, and (b) shall be treated as [shall be deemed to be] the income of the owner or beneficiary for that chargeable period, and (c) [shall not be deemed to be the income of any other person]

(2) This section does not have effect in relation to a sale or transfer if the proceeds of the sale or transfer are chargeable to tax. [For the purposes of subsection (1) above, in the case of a sale or other realisation the proceeds of which are chargeable to tax by virtue of section 18(3B) the interest so deemed to be the income of the owner or beneficiary shall be deemed to be equal in amount to the amount of those proceeds.] (2AA) [This section does not have effect for the purposes of Chapter 2 of Part 4 of the Finance Act 1996 (loan relationships).]

(3) The proceeds of any subsequent sale or other realisation of the right to receive the distribution shall not, for any of the purposes of the Tax Acts, be regarded as the income of the seller or the person on whose behalf the right is otherwise realised. [Nothing in subsection (1) above shall affect any provision of this Act authorising or requiring the deduction of income tax – (a) from any interest which, under that subsection, is deemed to be the income of the owner or beneficiary, or (b) from the proceeds of any subsequent sale or other realisation of the right to receive that interest; but the proceeds of any such subsequent sale or other realisation shall not, for any of the purposes of the Tax Acts, be deemed to be the income of the seller or the person on whose behalf the right is otherwise realised.] (7) In this section – ‘distribution’, in relation to shares in a company, - (a) has the same meaning as it has in the Corporation Tax Acts (see section 209), but (b) also includes any amount that would a distribution if the company paying it were resident in the United Kingdom; ‘shares’ means shares in a company.” [In this section – ‘interest’ includes dividends, annuities and shares of annuities, and ‘securities’ includes stocks and shares.”]

[572]As highlighted above, the main changes made in 2005 to the preceding version to produce the applicable version, as explained in the explanatory notes issued at that time, were as follows:(1) The explanatory notes introduce the provisions which include the relevant changes by stating that they “close a number of loopholes and block a number of avoidance schemes disclosed under part 7 Finance Act 2004 and elsewhere... the main categories of effective schemes are ones which convert interest-type income into a capital gain or an untaxed receipt.”

(2) The provisions originally related to both sales of dividends/distribution and interest coupons. However, the extension of the loan relationship rules applicable to securities meant that it was no longer necessary for s 730 to apply to sales of interest on securities (as set out at notes 26 and 27 of the explanatory notes). So s 730 was changed to confine its scope in the applicable version to sales of income distributions on shares/stock. Hence the amendment of the relevant definitions.

(3) No real explanation is given in the notes for the removal of the previous s 730(1)(c) which provided that the income distribution which was deemed to be taxable under s 730(1) was not to be deemed to be the income of any other person. It was merely stated, at note 17, that the relevant provision “removes the non-application of any tax charge on the recipient of the actual dividend.”

(4) A new s 730(3) was introduced which, at 20 of the notes, was stated to be needed to “ensure that there is no double taxation on the sale of coupons”.

(5) The changes from the previous wording in various places from “shall be deemed to be” to “shall be treated as” were stated in the notes to be to modernise the language.

(6) The change to sub-s (2) was stated to be to reflect changes made by the Income Tax (Trading and Other Income) Act 2005 and reverse the order of priority as between other charges on the sale of coupons. The other charges referred to appear to be those under the dividend investment provisions which tax the proceeds of sale of dividends on non-UK investments in certain circumstances. Discussion on s 730 issue – submissions

[573]BNP did not dispute that the aim of an exercise of statutory construction is to discern what Parliament intended, but that intention has to be discerned from the words chosen and, in this case, the wording of s 730(3) is clear and unambiguous. The very broad wording clearly covers income received in the form of trading income which would otherwise be taxable as such. The provision, therefore, excludes the proceeds of the sale of the right to the dividends realised by the London branch from the scope of s 730(1).

[574]BNP said that there is no indication in s 730 that it is not to apply to financial traders. On the contrary, the very broad references in sub-s (3) to “any subsequent sale” of the dividend coupon and that the proceeds of any such sale are not to be regarded as the income of the person “for any of the purposes of the Tax Acts” suggest it applies to traders. The most obvious person who would fall within this is a trader and the most obvious head of charge is that on trading income. Indeed, leaving aside the provisions of s 18(3B) ICTA and Chapter 1 Part 4 ITTOIA 2005 (the “foreign dividend provisions”) to which HMRC referred, there does not appear to be any other relevant head of charge which could be in point.

[575]The appellant considered that this interpretation is supported by the House of Lords’ decision in Hughes. The issue in that case was whether trading receipts, which comprised interest income, were exempt from tax under provisions which allowed the Treasury to issue securities with a condition that interest thereon “shall not be liable to tax or super tax” and the rule made thereunder that “no tax shall be chargeable in respect of … [broadly, certain interest or dividends on certain securities belonging to non-UK residents]”. Lord Thankerton rejected, at [374], the argument that the relevant provisions merely exempted relevant interest paid to non-UK residents from the head of charge for interest “qua interest” (under schedule D case III) and did not exclude it from being taken into account as a trading receipt in the computation of trading profits (under schedule D case I).

[576]HMRC argued that, on a purposive approach, looking at the history of the provision, as an anti-avoidance provision introduced in response to Paget, and the limited intended effect of the changes made in 2005, it cannot be the case that Parliament intended to create the effect that BNP’s interpretation would have. It is clear from the language used in the preceding version, that it would not have had that effect. BNP’s approach is based on a literal and strained construction of the statutory language, without proper regard to the legislative history and purpose of the relevant provisions. Such an approach is outdated and unpersuasive in the light of the modern approach to statutory interpretation as, set out in Barclays Mercantile Business Finance Ltd v Mawson

[2005] 1 AC 684

and, most recently, in UBS AG v Revenue and Customs Commissioners & DB Group Services (UK) Ltd v Revenue and Customs Commissioners

[2016] UKSC 13

[2016] 1 WLR 1005

(see Lord Reed at [61]).

[577]HMRC referred to the limited effect the changes to the provisions were intended to have, as set out in the explanatory notes. They noted, in particular, that the notes described the purpose of the specific change to s 730(3) as being “to ensure that there was no double taxation on the sale of the coupons”. They asserted that it is clear that explanatory notes may be taken into account as an aid to construction from the case of R v Montila

[2004] 1 WLR 3141

where the House of Lords, at [35], approved what Lord Steyn said in Westminster City Council v National Asylum Support Service

[2002] UKHL 38

[578]In HMRC’s view it is clear that, as both the explanatory notes and statutory language reveal, s 730 is not addressing the situation where any proceeds fall within the computation of a trader’s trading profits. It is aimed at cases where a sale of the right to future dividends would, absent s 730(1), give rise to a capital receipt and where it is, therefore, appropriate to treat the full amount of the dividend comprised within the coupon as the income (not merely a trading receipt) of the relevant person (under s 730(1)). This is not necessary as regards a financial trader, for whom the proceeds of any sale of the distribution would constitute a trading receipt in any event. The provision in sub-s (2), that s 730 is not to apply where the proceeds of a sale or transfer are chargeable to tax, is broad enough to cover proceeds comprising receipts which are brought into account in computing a trader’s taxable income. As noted, the purpose of s 730(3) was to avoid double taxation in this situation but there is no such concern where the re-seller is a financial trader.

[579]The wording of the provision does not, as BNP asserted, indicate that sub-s (3) is intended to apply to traders; investors also purchase and subsequently sell financial instruments. On the contrary the “proceeds of sale” on the sale by a trader of a right to a distribution would not normally be regarded as “the income” of the trader but simply as a receipt that is taken into account in computing that income, after deduction of expenses. Accordingly, the inference is that sub-s (3) applies only where a subsequent sale is not by way of trade but the proceeds would otherwise be treated as taxable income arising to the seller, such as under the foreign dividend provisions. The function of s 730(3) is simply to prevent those charges arising.

[580]HMRC referred in particular to the use of the term “the income” of the subsequent seller as indicating that sub-s (3) simply does not apply where a person receives a trading receipt as opposed to pure income profit. HMRC noted that Parliament did not use the broader wording that the proceeds of any subsequent sale cannot be “chargeable to tax” (the language used in sub-s (2)). The principle that a trading receipt is not “income” is well established in the UK tax code; it is only the overall profit of a schedular computation which is treated as income under the Tax Acts (as Hoffmann J held in George Wimpey International Ltd v Rolfe (Inspector of Taxes)

[1989] STC 609

at 615). The distinction between income and income receipts has long been recognised in respect of coupons/dividends and financial traders. This was a point made by Lord Romer in the Paget case itself and also in Cenlon Finance Co Ltd v Ellwood (Inspector of Taxes)

[1962] AC 782

(per Lord Reid at 795).

[581]HMRC regarded the decision in Hughes as of no assistance to the appellant. They noted that the wording used in the relevant provisions was very different to that in sub-s (3); it stipulated either that the interest would not be “liable to tax” or that “no tax shall be chargeable” in respect of interest. Moreover, it was found that the legislation in issue clearly manifested an intention on Parliament’s part to exempt the interest from income tax (the purpose being to encourage non-residents to invest in gilts). Hughes, therefore, provides no help in the construction of this provision given the different wording and purpose of the legislation in issue.

[582]In HMRC’s view their interpretation is supported by the more recent authority of the Court of Appeal’s decision in Vojak (Inspector of Taxes) v Strand Options & Futures Ltd

[2003] EWCA Civ 1457

[2004] STC 64

(‘Strand Options’). In that case, the dispute was whether s 208 ICTA 1988 provided an exemption from corporation tax on chargeable gains. Section 208 stated:

“208 Except as otherwise provided by the Corporation Tax Acts, corporation tax shall not be chargeable on dividends and other distributions of a company resident in the United Kingdom, nor shall any such dividends or distributions be taken into account in computing income for corporation tax.”

[583]The dispute was whether the wording that corporation tax “shall not be chargeable on … distributions” applies only where the distribution is a pure income receipt or also where it is included as part of a chargeable gain. The court agreed with HMRC that the words “chargeable on … distributions” suggest a tax which is directly charged on dividends/distributions as such, rather than charged only indirectly when the income is included in and is part of the computation of a taxable amount. They said this was reinforced by the contrast with the second part of s 208, which refers specifically to dividends or distributions being “taken into account in computing income”. In a capital gains context, the distribution is not directly the subject of tax, but is one element taken into account in computing the chargeable gain. If it had been intended to exclude it from the capital gains computation, it would have been “more natural to do so by including a specific reference to chargeable gains in the second part of s 208…”

[584]In HMRC’s view, there is a strong analogy between the reasoning in Strand Options and their position in this case. As in that case, there is a distinction to be drawn between the taxability of the dividend/distribution as pure income and as one element of the overall computation of a trader’s profits. That the legislature has stated that the relevant proceeds shall not be regarded as “the income” of the relevant person and has not used the broader formula, that they are not to be taken into account in computing income, clearly indicates that the provision only applies to prevent them being taxable as income as such. It does not prevent such proceeds from being brought into account as one element in a trader’s overall trading income computation under schedule D case I. Had that been the intention, the legislature could have used the broader formula (as in the second part of s 208) but has not done so.

[585]HMRC also noted that, in Strand Options, the Court of Appeal rejected the taxpayer’s reliance on the Hughes case. At [19], Carnwath LJ said he did not find this a helpful authority in the present context as, as the Judge had said in the High Court, “the Court’s interpretation of completely different tax statutes, relating to different subject matter and dating from a completely different era, can only be, at best, of marginal assistance in the interpretation of TA 1988 s 208.”

[586]BNP countered that the history of s 730 cannot be of material assistance given it has undergone substantial changes, including major changes from the preceding version to the applicable version. Moreover, whilst s 730(1) is an anti-avoidance provision (which in effect deals with future Miss Pagets), sub-s (3) itself is concerned with avoiding double taxation as regards a person who purchases the relevant distribution rights and sells them on.

[587]BNP noted that, on HMRC’s interpretation, s 730(3) would be limited to apply solely to non-dealers in coupons who purchase and subsequently sell rights to distributions on shares in foreign companies by way of investment (on the basis it applies only where the foreign dividend provisions apply). If that was really what the draftsman of sub-s (3) intended, the provision could simply have said so. The draftsman must have been aware of the foreign dividend provisions as they were specifically referenced in the preceding version (see the previous sub-s (2)). It appears, therefore, that the draftsman chose to express sub-s (3) in the widest possible terms.

[588]BNP asserted that HMRC’s position, that sub-s (2) is broad enough to exclude traders from the scope of s 730(1) (on the basis that in their hands “the proceeds of the sale or transfer are chargeable to tax”) but that such traders are not within sub-s (3), is inconsistent. Both sub-sections concern the taxability of the proceeds of sale; if traders come within the ambit of sub-s 2, they must logically also fall within sub-s (3).

[589]In BNP’s view, if it is accepted that s 730(3) applies in principle to financial traders, it must follow that that the relevant proceeds of sale are not to be regarded as a taxable trading receipt in the trader’s hands. Such proceeds of sale can never be “pure income profit” in the hands of a trader because the purchase price and other expenses are set against the proceeds. On HMRC’s interpretation, therefore, s 730(3) would never in practice apply which cannot be the intention. There is nothing in the language of the provision to suggest that, in this context, “income” should be limited to “pure income profit”. Indeed, the use of the word “proceeds” makes it clear that the provision is concerned with receipts rather than profits.

[590]In BNP’s view HMRC’s interpretation involves speculating as to the intention behind the provisions or assuming what the intention is and then seeking to strain or rewrite the words of the statute to make the two align. However, Chadwick LJ warned against that approach in the Court of Appeal decision in Frankland v IRC

[1997] STC 1450

at page 1464 (in the context of inheritance tax but the principle is of universal application). He said that whilst legislation must be construed in its statutory context and with due regard to the purpose which the legislature may be taken to have been seeking to achieve:

“that purpose must, I think, be identified in the legislation itself and in any other relevant and admissible material. It is not permissible to speculate, a priori, as to what the legislature must or might have intended, and then strain the statutory language used in order to give effect to that presumed purpose.”

[591]HMRC responded that BNP’s interpretation gives rise to absurd results. The tribunal must adopt an approach that avoids the absurdity that BNP’s interpretation produces as was set out by Neuberger J (as he then was) in Jenks v Dickinson (Inspector of Taxes)

[1997] STC 853

at 870 and as indicated in the cases of Luke v IRC

[1963] AC 557

(at 577), 40 TC 630 at 646 and O’Rourke (Inspector of Taxes) v Binks

[1992] STC 703

at 579–580, 40 TC 630 at 648. Unlike some of the cases cited by Neuberger J, in this case the tribunal does not have to use a strained interpretation. Rather, it merely needs to give proper consideration to the words that Parliament has used in s 730(3), having regard to the legislative purpose of both the original provision and the amending provision. HMRC are not starting with an a priori assumption as to what the provisions mean. They are simply looking at the statutory context and at the actual wording of the provision. On that basis, it is clear that “the income” means effectively pure income profit. When the Parliamentary draftsman wants such an amount to be excluded also from being taken into account in computing income or computing gains, there is a well-established form of wording which is normally followed (as in s 208) but that is not used here.

[592]BNP said that, even if there were such an anomaly as HMRC suggest, that cannot justify HMRC’s construction of s 730(3), the wording of which is absolutely clear. If, on a proper construction of those words, their application leads to a particular result, then that result has to be respected even if it could give rise to the outcomes which HMRC assert are anomalous unintended results. There are occasions where the legislation simply gives rise to such results, particularly, as regards overlapping codes of complexity, such as, the interaction of s 730 with the rules about deduction of expenditure. This was recognised by Henderson J in D'Arcy at [47] and recently in Bowring and another v Revenue and Customs Commissioners

[2016] STC 816

where Mr Justice Barling said one cannot, and should not “... strain the ... clear meaning and effect of the legislation, to the extent which would be necessary to produce the outcome for which [HMRC] argued.”

[593]The absurd results which HMRC referred to are that, in their view, on BNP’s approach, s 730(3) would apply in circumstances where there is no risk of double taxation and would open the door to multiple avoidance of tax. On BNP’s interpretation, a share dealer who is the first to sell the coupon (but retains the shares), would realise taxable income equal to the amount of the distribution (under s 730(1)) but the second and any subsequent share dealers who acquire and sell the coupon would enjoy a fiscal windfall. Each such trader could deduct the cost of acquisition but, on the appellant’s view of s 730(3), could exclude the receipt from the trading computation, thereby almost guaranteeing the realisation of a trading loss which bears no reality to the financial outcome for the trader. These are manifestly absurd results which cannot have been the intention of Parliament.

[594]BNP responded that the position HMRC set out is not, in fact, generally the result which follows from its interpretation. In a typical case, whilst a financial trader would obtain a deduction for the purchase price paid for the acquisition of the dividend right for the purposes of his trading computation, he would be liable to tax on a chargeable gain computed by reference to the proceeds of sale. That is because s 730(3) does not exclude the proceeds from being taxed as a chargeable gain and the trader would not obtain a deduction in the capital gains computation for the purchase price (under s 37(1) and s 39(1) TCGA.) The London branch only escapes this charge because it is not within the scope of corporation tax on capital gains as regards the sale of the right to dividends.

[595]HMRC saw two difficulties with BNP’s analysis.(1) There is no rational explanation why Parliament would have intended to create a hybrid asset, as is the result of BNP’s analysis. Particularly in relation to individuals, income tax and capital gains tax have materially different rules for the recognition of receipts and expenditure, the availability of reliefs, the use of losses and the rates at which tax is charged. A policy of allocating receipts and expenditure relating to the same asset to different taxes would be guaranteed to result in complexity and anomaly.

(2) On BNP’s analysis, a non-resident carrying on a trade through a permanent establishment in the UK could deduct the cost of a right to dividends in computing the trading profits of the permanent establishment but would fall outside the scope of any charge to tax on the proceeds of sale altogether (as it would not be within the scope of tax on chargeable gains). There is no rational reason why Parliament would want to create such a situation.

[596]BNP asserted that HMRC’s construction would result in the very double taxation which the explanatory notes state s 730(3) is intended to avoid. A “subsequent” trader would be liable to tax on the proceeds of sale notwithstanding the proceeds reflect the value of the dividend on which the original seller is taxed. The final purchaser who actually receives the dividend would also be subject to income tax on the income (assuming he acquired the right on investment account) because of the repeal of the previous s 730(1)(c). Excluding the proceeds of sale realised by the trader from tax in such circumstances has precisely the intended effect of avoiding double taxation. In this case, BNP Lux is not potentially taxable under s 730(1) in any event because it is not within the scope of UK tax on the receipt (under s 151 of the Finance Act 2003). But that does not make any difference to the proper interpretation of the provision (and HMRC do not suggest that is the case).

[597]HMRC noted that the “subsequent” trader would only be taxable on the real surplus or loss generated from the purchase and re-sale of the right to the dividend (and not the amount of the dividend itself); the profit/loss would be computed in the usual way with the financial trader obtaining a tax deduction for the price it paid to acquire the dividend rights. They agreed that, following the repeal of s 730(1)(c), the dividend could be taxable as income in the hands of the actual recipient in accordance with normal principles. However, in their view, the incidence of double taxation is likely to be more theoretical than real; in most circumstances, either the transferor or the transferee (or both of them) are likely to be outside the charge to tax in respect of such dividends. As the House of Lords’ decision in R v Dimsey

[2001] UKHL 46

[2001] STC 1520

(at [56] to [60]) shows, in the context of an anti-avoidance provision, such largely theoretical possibilities should not distort the construction of the provision. HMRC concluded that in any event, BNP’s construction of sub-s (3) does nothing to remedy such possibilities. It applies only to the proceeds from a subsequent sale or realisation of the right to the dividend, it does not impact on the taxation treatment of the recipient of the actual dividend. Discussion and conclusion

[598]There is no dispute that under the modern approach to the interpretation of tax legislation, as is well established in case law, we must apply a purposive approach to the interpretation of s 730(3). In our view, looking at s 730(3) in the overall context of s 730, the wording used indicates that it does not apply to exempt a trader, who makes a subsequent sale of a right to a distribution, from tax on any trading profit realised on that sale as computed under the normal rules.

[599]As noted and as was common ground, s 730 was introduced to prevent investors realising a capital profit on the sale of an income stream. Although the section has been changed a number of times over the years, there is no reason to think that the rationale for its existence changed. The explanatory notes issued in relation to the applicable version cite that as the reason for the provision.

[600]Section 730(1) provides that where the owner sells or transfers the right to receive any distribution in respect of shares without a sale/transfer of the shares then “that distribution … shall be treated as the income of the owner [or of the beneficial owner of the income]”. This is clearly aimed at rendering the distribution itself taxable as such in the hands of the owner. Section 730(2) provides that s 730 does not have effect in relation to a sale or transfer of the right to a distribution, without a sale/transfer of the shares, if the proceeds are chargeable to tax. Where the owner is a trader the proceeds of the sale of the right are chargeable to tax in the trader’s hands in that they are brought into account in the computation of his trading profits in the usual way. It is clear, therefore, that the combined effect of these sub-sections is that a trader is not within the scope of a charge under s 730(1). That accords with the rationale of preventing the conversion of what would otherwise be income receipts into capital returns.

[601]To recap, s 730(3) provides that “the proceeds of any subsequent sale or other realisation of the right to receive the distribution shall not, for any of the purposes of the Tax Acts, be regarded as the income of the seller or the person on whose behalf the right is otherwise realised”.

[602]There was similar wording in the preceding version although it was more limited in scope. In summary, the previous provision stated that nothing in s 730(1) prevented the deduction of income tax from interest which was deemed under that section to be the income of the owner or from the proceeds of any subsequent sale of such a right to interest but that such proceeds shall not, “for any of the purposes of the Tax Acts, be deemed to be the income of [the relevant party]”. The relevant wording only applied, therefore, to the proceeds of a subsequent sale of interest income. The deeming wording indicates that such proceeds, which would otherwise be taxable in the hands of the relevant party (such as where they were in any event taxable as trading income), would remain so taxable. The concern appeared to be that such proceeds would not be deemed to be taxable as interest income as such. In the preceding version, therefore, s 730(3) was not providing a trader, who buys and sells a right to income, with a blanket exemption from tax on the proceeds of sale.

[603]The explanatory notes state that the aim of s 730(3) in the applicable version is to ensure that there is “no double taxation on the sale of the coupons”. We consider that, from the wording of the provision itself and in the context of the clear meaning of the other provisions of s 730(1), the better interpretation of this provision is that it is confined to preventing the proceeds of sale from being taxed as though they are “the income”, in the sense of pure income profit, such as a distribution, of the relevant recipient. We accept that, as argued by HMRC, due to the operation of the schedular system as in place at the time, there is a distinction between a trading receipt which is brought into account as an element in computing a trader’s overall trading income and a pure income receipt such as a distribution or interest income. The reference to the proceeds falling within s 730(3) not being regarded as “the income” of the relevant person implies that it is only the prevention of a charge on the relevant proceeds as pure income which is covered.

[604]We do not consider that this is a strained interpretation of the wording used. It is simply the outcome of applying a purposive approach in this case. The interpretation we adopt leads to the result that a trader is largely unaffected by the s 730 rules and simply taxed on the resulting profit from the buying and selling of the dividend rights in the usual way. This is in accordance with the aim behind the legislation of preventing tax avoidance by investors in effect converting income into capital. It accords with the fact that a trader is clearly excluded from any charge under s 730(1) on the basis that the proceeds are taxable in the trader’s hands in any event. On the other hand, BNP’s approach produces the manifestly absurd result that a trader who first sells the dividend rights is outside the scope of the main charging provision, such that he is taxed in the normal way on any resulting profit, but any subsequent trader who buys and sells the right would be wholly outside the scope of any tax charge whatsoever. Nor do we see that the provisions can be said, on BNP’s interpretation, to operate to prevent a double charge to tax in the intended way where, for example, the subsequent trader buys the right from an investor who is within the scope of s 730(1). It remains the case that it would be wholly out of kilter with the overall operation of these provisions for a subsequent trader to escape any tax charge on the proceeds of sale it realises.

[605]We find it very difficult to see that the legislature intended the meaning of the provision which BNP argued for on the basis that the otherwise manifestly absurd tax effects would be “made good”, as BNP asserted, through the combined effect of the income and capital gains rules. Such a hybrid approach would be highly unusual and of itself gives rise, as HMRC noted, to a number of problems.

[606]We appreciate that this interpretation means that, in practice, the application of s 730(3) is limited to cases where the dividend income provisions apply which is a narrow range of cases. However, we do not see that of itself affects matters. Section 730 is a relatively narrow anti-avoidance provision applicable in limited circumstances. That there are only limited situations in which there could be a double tax charge on the sales proceeds as “income” is not perhaps surprising.

[607]In our view, the Hughes case does not add to the debate given the different wording in issue and the purpose of those rules. Similarly Strand Securities, as a decision on the interpretation of a different set of provisions on the particular wording used, can only be of limited assistance. We have based our decision on the interpretation of the wording used in this provision as set out above. Conclusion

[608]For all the reasons set out above, we have concluded that.(1) The transaction comprising the purchase and sale of the right to the dividends by the London branch was not undertaken by it in the course of its financial trade.

(2) In any event, the price paid by the London branch for the acquisition of the right to the dividends was not incurred wholly or exclusively for the purposes of the London branch’s trade but at least in part (if not solely) for the purpose of realising the s 730 benefit.

(3) HMRC are entitled to raise the s 730(3) argument.

(4) Section 730(3) does not operate to exclude the proceeds of sale realised by the London branch on the sale of the right to the dividends to ALIL from being brought into account in the computation of its trading income for corporation tax purposes.

[609]The appeal is accordingly dismissed.

[610]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. GREG SINFIELD HARRIET MORGAN TRIBUNAL JUDGE TRIBUNAL JUDGE RELEASE DATE: 12 JUNE 2017 APPENDIX REFERENCE LIST – NAMES AND ABBREVIATIONS Name Description A&L Alliance & Leicester plc, the parent of ALIL ALIL Alliance & Leicester Investments Limited, a subsidiary of A&L Bayfield, Stuart A member of OF Berg, Eric A member of the BNP Lux tax team BNP BNP Paribas SA, parent company of the BNP Group BNP Group BNP Paribas Group which provides banking and financial services in various countries BNP Lux BNP Luxembourg SA, a subsidiary of BNP BNP UK BNP PUK Holding Limited, a subsidiary of BNP Delafontaine, Christophe Global Head of OF and Chair of the TAC based in Paris Demon, François Head of GF GF Gestion Financiere, a division of BNP Group’s finance department HIL Harewood Investments No. 5 Limited HMRC The Respondents, the Commissioners for Her Majesty’s Revenue and Customs London branch The London branch of BNP Majchrzak Gilot, Stephanie Head of the BNP Lux structured transactions team OF Optimisation Finance, a team in the London branch which put together the transaction Peters, James A member of OF Pouliguen, Patrice Group Head of tax of the BNP Group Read, Oliver A member of OF Robinson, Neil Head of OF in the UK Scholes, Peter Head of UK tax Stanton, Marcus A banking consultant and expert witness for HMRC Trifiletti, Robert A member of BNP’s UK accounting team TAC Transactions Approval Committee TCC Tax Coordination Committee of the BNP Group Williams, Nick A member of OF

BNP Paribas SA (London Branch) v Revenue and Customs (INCOME TAX/CORPORATION TAX : Anti-avoidance) [2017] UKFTT 487 (TC)

(a) It was described in the heading as:

Cited in 2 later judgments