The Commissioners for HM Revenue and Customs v Smith and Nephew Overseas Ltd and Others: [2018] UKUT 0393 (TCC) [2018] UKUT 0393 (TCC)

UPPER TRIBUNAL

TAX AND CHANCERY CHAMBER

[2018] UKUT 0393 (TCC)

Case No UT/2017/0075

Between

The Commissioners for HM Revenue and CustomsAppellant

Smith and Nephew Overseas Ltd and OthersRespondent

Before

MR JUSTICE FANCOURT

JUDGE THOMAS SCOTT

Julian Ghosh QC, Jonathan Bremner QC and Charles Bradley (instructed by Johnson Allen Tax) for Respondents

Date 29 November 2018

Category: Tax

[37]We were also concerned with Mr Chopping’s reliance on KPMG’s 2007-08 manual in his Supplemental Report to counter the statement in Mr Hogarth’s Report in relation to the Deloitte manual. Although it would appear that his failure to refer to the 2015-16 KPMG manual, of which he was not aware, was, to adopt the language of the penalty 20 provisions, careless rather than deliberate, we would have expected an expert witness to have been aware of the latest guidance and ensure it was brought to our attention.

[38]We therefore conclude that by adopting the Foreign Operation method the accounts of SN Overseas, TP and SN Finance do comply 25 with UK GAAP. We find support for our conclusion not only from the evidence of Mr Hogarth but also the manuals of Deloitte, PwC and KPMG and the fact that on 13 August 2009, EY signed the audit opinion in respect of SN Overseas and TP to the effect that these companies accounts gave a True and Fair view (see paragraph 5(55), above).” 30 HMRC’s arguments HMRC submit that the FTT “conspicuously failed to engage” with the relevant questions, instead determining the issue on the flawed bases of a “bald assertion” that they preferred Mr Hogarth’s evidence to that of Mr Chopping, the supposed support in the three Manuals and the errant reliance on the EY audit opinion. HMRC’s 35 skeleton argument devotes some 25 pages to submissions on this issue. Unpicking those 25 pages, HMRC’s arguments are as follows:(1) The FTT’s conclusions as regards the evidence were conclusions that no reasonable tribunal, acting judicially and properly instructed as to the relevant law, could have reached: see Edwards v Bairstow

[1956] AC 14

at 40 pages 29 and 36.

(2) The FTT erred in law in failing to give proper reasons for its decision: Flannery v Halifax Estate Agencies Ltd

[1999] 1 WLR 377

. In so far as reasons were given, those reasons were unsupported by the evidence, irrelevant and based on a misunderstanding of both the issues and HMRC’s 45 case.

(3) The result produced by the FO Method in this case (substantial exchange losses) was plainly inappropriate because “measured in sterling the 18 companies had no economic exposure to foreign exchange risk and suffered no economic loss throughout 2007 and 2008”.

(4) Mr Hogarth’s opinion that it was more appropriate to use the FO Method was wrong in the circumstances measured against all four FRS 18 criteria. By contrast, Mr Chopping’s opinion that the SR Method was the most 5 appropriate was correct by reference to all four criteria.

(5) It was the duty of the Tribunal to consider and reach its own conclusions on the proper accounting treatment and to give detailed reasons for those conclusions. It was an error of law for the FTT to have abrogated that responsibility by the adoption of “an unwarranted global preference” for Mr 10 Hogarth over Mr Chopping.

(6) Where the FTT prefers one expert over another, it must give cogent reasons for that preference, and the FTT’s failure to do so was an error of law: Flannery v Halifax.

(7) The deficiencies of the Decision are plain when compared to the detailed 15 reasoning of the FTT in Ball UK Holdings Ltd v HMRC

[2017] UKFTT 457

(TC).

(8) The FTT erred in law in relying expressly on the contents of the Manuals, because the judgment as to the most appropriate accounting treatment depended on all the circumstances and so could not be addressed 20 in the abstract.

(9) The FTT erred in law in relying on approval of the accounts by the company’s auditors. Such approval did not necessarily establish that the accounts were GAAP compliant: see Ball UK Holdings.

(10) The FTT’s analysis of each of the four FRS 18 criteria was wrong in 25 law. S&N’s arguments For the S&N Companies, Mr Bremner QC and Mr Ghosh QC made the following submissions: (1) HMRC’s appeal was fundamentally based on challenges to the FTT’s 30 findings of fact, over which the Upper Tribunal had no jurisdiction. (2) The Upper Tribunal should exercise caution in asserting jurisdiction on the basis of challenges to findings of fact dressed up as Edwards v Bairstow errors of law. See, in particular, the judgment in Georgiou v Customs and Excise Commissioners

[1996] STC 463

at 476. 35 (3) That caution should be even greater where, as here, the FTT has made findings of fact based on expert evidence. (4) The FTT’s findings in relation to Mr Chopping’s role as an expert witness were par excellence findings of fact, and disclosed no error of law. (5) Before the FTT, HMRC needed to establish not only that the SR Method 40 would have been an acceptable method to adopt, but that the adoption by the S&N companies of the FO Method was positively non-compliant with 19 GAAP. For several reasons, the FTT could not properly have reached this conclusion. (6) The submissions made in HMRC’s skeleton argument reduced on analysis to the complaint that the FTT’s assessment of the evidence was not the same as HMRC’s. 5 (7) The FTT plainly did give reasons for their decision. Applying Flannery properly, HMRC could readily tell from the Decision why they had lost on this issue. (8) In relation to the Manuals and to EY’s audit opinion, it was proper for the FTT to take these into account in reaching their decision, and evident 10 that they treated neither factor as determinative. (9) HMRC’s argument that the FO Method was plainly inappropriate given the absence of exposure to foreign exchange risk was flawed, because it led to the conclusion that that method could never be an appropriate method to account for a change in functional currency, and there had been no evidence 15 to that effect. Discussion Section 11(1) of the Tribunals, Courts and Enforcement Act 2007 (“TCEA”) provides that a party to a case before the FTT only has a right of appeal to the Upper Tribunal on a point of law arising from the FTT’s decision. There cannot be an appeal 20 on a pure question of fact which is decided by the FTT. However, a tribunal may arrive at a finding of fact in a way which discloses an error of law. That is clear from Edwards v Bairstow in which Lord Simonds referred to making a finding, without any evidence or upon a view of the facts which could not be supported, as involving an error of law: see page 29. In the same case, Lord Radcliffe, at page 36, regarded cases 25 where there was no evidence to support a finding or where the evidence contradicted the finding or where the only reasonable conclusion contradicted the finding, as cases involving errors of law.

[35]In relation to an appeal which is said to involve a point of law of the kind identified in Edwards v Bairstow, we are mindful of what was said by Evans LJ in Georgiou v 30 Customs and Excise Commissioners

[1996] STC 463

at 476, as follows:

“It is right, in my judgment, to strike two cautionary notes at this stage. There is a well-recognised need for caution in permitting challenges to findings of fact on the ground that they raise this kind of question of law. That is well seen in arbitration cases and in many others. It is all too easy 35 for a so-called question of law to become no more than a disguised attack on findings of fact which must be accepted by the courts. As this case demonstrates, it is all too easy for the appeals procedure to the High Court to be misused in this way. Secondly, the nature of the factual inquiry which an appellate court can and does undertake in a proper case 40 is essentially different from the decision-making process which is undertaken by the tribunal of fact. The question is not, has the party upon whom rests the burden of proof established on the balance of probabilities the facts upon which he relies, but, was there evidence before the tribunal which was sufficient to support the finding which it 45 20 made? In other words, was the finding one which the tribunal was entitled to make? Clearly, if there was no evidence, or the evidence was to the contrary effect, the tribunal was not so entitled.”

He continued:

“... For a question of law to arise in the circumstances, the appellant must 5 first identify the finding which is challenged; secondly, show that it is significant in relation to the conclusion; thirdly, identify the evidence, if any, which was relevant to that finding; and fourthly, show that that finding, on the basis of that evidence, was one which the tribunal was not entitled to make. What is not permitted, in my view, is a roving 10 selection of evidence coupled with a general assertion that the tribunal’s conclusion was against the weight of the evidence and was therefore wrong. A failure to appreciate what is the correct approach accounts for much of the time and expense that was occasioned by this appeal to the High Court.” 15 We consider first HMRC’s assertion that the FTT failed to give reasons, or adequate reasons, for its decision. HMRC’s argument that, by contrast with Ball UK Holdings, the FTT failed to reach independent conclusions and explain those conclusions in detail is considered in that context. Ball UK Holdings (the appeal in which is listed to be heard before this Tribunal) 20 does indeed demonstrate more detailed and granular reasoning than the Decision. It is likely that the FTT in Ball was properly mindful of the importance of setting out detailed reasons for finding that the accounts in that case were not GAAP compliant notwithstanding the views of two of the three experts and an audit opinion as to a true and fair view. In any event, we reject the proposition that in reaching a judgment based 25 to a material degree on expert evidence the FTT makes an error of law if it does not perform for itself the same step-by-step analysis as the expert or experts, setting out in detail at every step and in exhaustive detail what Mr Rivett referred to as “its workings”

. The duty to give reasons was considered by the Court of Appeal in the context of 30 a judge preferring one expert over another in Flannery v Halifax. Henry LJ, handing down the judgment of the court, affirmed the existence of a general duty to give reasons, but stated that it was “not a useful task to attempt to make absolute rules as to the requirement for the judge to give reasons”:

[1999] 1 WLR 377

at page 381. As emphasised by Henry LJ, fairness dictates that the parties, especially the losing party, 35 should be left in no doubt as to why they have won or lost; how that requirement is fulfilled depends on all the circumstances. In this case, it would have been clear to HMRC why they lost on this issue. The FTT considered the expert evidence, the Manuals and, by reference to the expert evidence, the four FRS criteria. It then set out its conclusions, with reasons, at 40 paragraphs [36] to [38]. HMRC patently did not agree with the conclusions or reasons, but that is a different matter. Reading the Decision as a whole, we consider it is clear that on this issue the FTT was adopting the expert evidence of Mr Hogarth. It was fully entitled to do so. It is also clear that the FTT gave reasons for its decision, referring not only to the expert 45 21 evidence but also to the “support” they found for their conclusion in the Manuals and noting the EY audit opinion. We do not accept that those reasons were based on inappropriate factors or evidence, or any misunderstanding of the issue. The question of whether the judgment formed by the directors of the S&N Companies was in accordance with GAAP was 5 self-evidently an issue apt to be considered with the benefit of expert evidence. In establishing evidence of practice, the Manuals were relevant, and the FTT rightly identified them as “not authoritative”. The audit opinion was relevant but (as was made clear in Ball) not determinative, and the FTT gave no indication that it considered otherwise. We note that HMRC did not identify other evidence or criteria 10 which the FTT should have taken into account, and, of course, HMRC had the opportunity so to argue before the FTT. We turn next to HMRC’s assertion that the FTT’s preference for the expert evidence of Mr Hogarth was based on a “bald assertion” and unwarranted on the evidence. 15 The FTT describes its preference for Mr Hogarth’s expert evidence at [36] and [37] of the Decision. It is made explicit in these passages that the FTT had in the course of the hearing formed reservations in relation to the objectivity and expertise of Mr Chopping. Both of these qualities were critical to his role before the tribunal as an expert witness. So, far from being a bald assertion, the FTT described and justified 20 the reasons for its preference. Were those reasons, as HMRC must establish, based on irrelevant factors or no evidence, or perverse? We were taken by Mr Rivett to a considerable number of instances in the transcript of the FTT hearing, which he submitted demonstrated HMRC’s case. Having considered those examples, we find HMRC’s submission to 25 be ill-founded. Indeed, the detailed exchanges between the FTT and Mr Chopping in our opinion tended to show that the FTT’s concerns were not unjustified or unfounded. HMRC’s case in relation to the expert evidence falls well short of establishing an Edwards v Bairstow error of law. We also place significant weight on the fact that the FTT, which was a specialist tribunal, had the benefit of hearing, observing and 30 questioning the two experts over an extended period. Both this Tribunal and the Court of Appeal emphasised the importance of that factor (again in relation to the FTT’s assessment of expert evidence as to GAAP compliance) in GDF Suez. HMRC submit that the FTT erred in “relying expressly” on the contents of the Manuals and on the approval of the accounts by the auditors. 35 That submission does not withstand scrutiny. We were taken through the relevant extracts from the Manuals, and from this we have concluded that the summary at [29] of the Decision was fair and reasonable. The FTT endorsed the view that while the Manuals were “not authoritative” they did describe what the relevant firms saw in practice: [30]. In “finding support” for their conclusion based on the expert evidence, 40 the FTT were clearly not treating the Manuals as determinative. The same is true of the reference to the EY audit opinion. An audit opinion does not prove GAAP compliance (see Ball) but since it is evidence of a professional opinion as to the true 22 and fair view criterion it is relevant. The FTT was clearly correct in not ignoring or discounting as irrelevant the Manuals and the audit opinion. HMRC’s submission in relation to the four FRS 18 criteria proved on consideration to be an instance of the “roving selection of evidence coupled with a general assertion that the tribunal’s conclusion was against the weight of the evidence 5 and was therefore wrong” identified in Georgiou. The FTT considered the expert evidence in relation to all four criteria and set out the reasons for its conclusions. Mr Rivett argued that HMRC viewed reliability as the primary criterion whereas the S&N Companies wrongly viewed the primary criterion as relevance. However, that is not the issue. HMRC must establish that the FTT’s reasons or conclusions in relation to 10 FRS 18 involved an error of law, and they have failed to do so. HMRC put forward a separate argument on this issue that the use of the FO Method was “plainly inappropriate” because of the absence of foreign exchange exposure for the S&N Companies as a result of the change in functional currency. We also consider this point in relation to the remaining two issues in the appeal. However, 15 the issue before the FTT in assessing GAAP compliance was whether the accounting method chosen by the directors of each company was “in accordance with GAAP”, not whether it was the best or a better method of recording foreign exchange exposure. The question of foreign exchange exposure was addressed in the expert evidence and the FTT refers explicitly to Mr Chopping’s views as to the relevance of the exposure 20 point: see [32], [33], [35]. More generally, the FTT was well aware of HMRC’s arguments as to the exposure point in relation to the “exchange loss” issue: see [42]. The FTT took the point into account and addressed itself to the correct question in relation to section 85A. It made no error of law. We deal finally with HMRC’s argument that the FTT misdirected itself in law 25 because it did not understand that in stipulating the use of “the most appropriate” method, FRS 18 required a binary analysis as between the two available methods, as only one of them could ever be “the most appropriate”. We consider this to be an unduly narrow reading of the relevant accounting standard. The starting point is the (agreed) absence of a specific accounting standard 30 at the relevant time. Had there been one, the issue would indeed have been whether that single standard had been properly adopted. In the event, the directors were required by FRS 18 to formulate an accounting policy, and were directed to take into account four criteria in that calculation, with the explicit objective of satisfying the true and fair view requirement. It seems to us, particularly in view of the conflicting 35 accounting evidence, that because this was a multi-factorial question of judgment and balancing several factors, more than one view might reasonably be formed as to which method was “most appropriate”. HMRC did not prove before the FTT or this Tribunal that only the SR Method was GAAP compliant, or that no reasonable tribunal could have reached the conclusion that the FO method was GAAP compliant. 40 Even if we are wrong on that issue, the FTT in this case in fact approached and answered the question as formulated by HMRC. It did not approach the issue by a consideration of what decision the directors could reasonably have made, and whether the method chosen was GAAP compliant, but in effect by asking itself which method 23 was the most appropriate and preferring Mr Hogarth’s evidence that in the circumstances the FO Method was more (and therefore “the most”) appropriate. We reach the clear conclusion that HMRC’s appeal on the first issue fails. Second issue: exchange loss HMRC appeal against the FTT’s decision that the claimed exchange differences 5 gave rise to “exchange losses”. The issue Even if the exchange differences shown in the relevant accounts were brought into account in accordance with GAAP, they would only be treated by section 84A as loan relationship debits if they were “exchange losses”. HMRC contend that the word 10 “loss” must be interpreted in the statutory context as applying only to a “real economic” loss which is actually suffered by a company. In this case, they argue, the S&N Companies showed purely arithmetical differences in their accounts, but had no exchange rate exposure and realised no currency losses. As a result, no loss, and therefore no exchange loss, arose. 15 The FTT Decision The FTT set out the competing arguments of the parties at [40] to [43] of the Decision. It concluded that Mr Ghosh’s submissions for the S&N Companies were correct, stating as follows: “44. Given that an exchange loss is the comparison at different times of 20 the expression in one currency of the valuation put by the company in another currency in relation to an asset we find that it must, in essence and notwithstanding Mr Rivett’s submissions to the contrary, be an arithmetical exercise. We also note that the legislation does not refer to there being any exposure to exchange rates between the two dates for 25 there to be such a loss, just a comparison at different times in one currency and then in another.

[45]Such a comparison at different dates is not a novel concept. In The Spanish Prospecting Company Limited

[1911] 1 Ch 92

Fletcher Moulton LJ considered the legal meaning of the word “profits” saying, at 98: 30 ““Profits” implies a comparison between the state of a business at two specific dates usually separated by an interval of a year. The fundamental meaning is the amount of a gain made by the business during the year. This can only be ascertained by a comparison of the assets of the business at the two dates.” 35

[46]Therefore, as there was a fall in the value of the assets, the Intercompany Receivables, in the present case at the second date 31 December 2008 when compared with their value as at 31 December 2007 as stated using the Foreign Operations method in what we have found to be UK GAAP compliant accounts it must follow that the 40 Exchange Differences are exchange losses within s 103.” 24 Arguments of the parties HMRC submitted that in context a “loss” arose only when there was an actual or economic loss. The word “loss” must be given its ordinary meaning in the general context of the statute: see R v Environment Secretary ex parte Spath Holme Ltd

[2001] 2 AC 349

and 379b. In Stagecoach Group plc and another v HMRC

[2016] UKFTT 5

120 (TC) in the context of the loan relationship rules, the FTT focussed on the existence or absence of an actual loss. In Union Castle Mail Steamship Co Ltd v HMRC

[2016] UKFTT 526 (TC)

in the context of the rules on derivative contracts the FTT held that the taxpayer had suffered no real loss because there had been no diminution in its resources. A similar approach should apply in this case, and the FTT 10 was wrong to distinguish Stagecoach and Union Castle. In this case, argued HMRC, the exchange differences were not actual or economic losses, and this was demonstrated by a number of points. The companies had no economic exposure to foreign exchange risk and suffered no economic loss throughout 2007 and 2008. The Group revaluation reserve arising from the change in 15 functional currency had been treated as non-distributable, showing that the exchange differences were not realised losses. SN Overseas had paid a dividend on 23 December 2009 on the basis that the differences were not realised losses. The additional share premium paid by SN Overseas and TP for the subsidiaries in the group reorganisation was effectively eliminated by capital reductions in 2013. Finally, the exchange 20 differences were not recognised in the Group’s consolidated accounts. In their skeleton argument, but not at the hearing, HMRC raised a separate argument that the losses were “actively engineered” by the S&N Companies, so as to exploit a known exchange rate movement and with the clear objective of generating exchange rate differences, using an accounting policy which would produce those 25 differences. For the S&N Companies, Mr Ghosh argued that HMRC should not be permitted by this Tribunal to raise the “actively engineered” argument, as it was not pleaded before the FTT and would clearly require evidence from the relevant directors and advisers in response. In any event, he submitted, the allegation was entirely unfounded 30 on the facts. In relation to the meaning of “exchange loss”, Mr Ghosh submitted that the definition in section 103(1A) was clear, and the FTT had correctly identified that it required nothing more than a comparative mathematical exercise. HMRC’s proposition that a loss resulting from the comparison must also be an “actual or 35 economic” loss derived no support from the words of section 103(1A) or the legislative context. That was not surprising in the context of a statutory scheme in which the relevant criterion was recognition of a profit or loss in the company’s STRGL. Moreover, this was consistent with the normal accounting practice of translating monetary assets denominated in a foreign currency at each balance sheet 40 date or transaction date. Mr Ghosh further submitted that, in any event, the exchange losses in this case did have an actual economic impact. In particular, because of the depreciation of sterling against the US dollar, SN Overseas and TP had to raise additional amounts of 25 capital (by way of share premium) in order to purchase shares in the group reorganisation. Discussion Mr Rivett did not pursue before us the argument that the exchange losses were “actively engineered”. He was right not to do so. We agree with Mr Ghosh that the 5 FTT was the proper forum in which to take this point, because it raises primarily issues of fact. Neither party sought to challenge the lengthy statement of agreed facts set out in the Decision and recorded at [9] above, and we take that statement as fairly recording the background to and reasons for the relevant events in this appeal. The question raised by this issue is one of statutory construction. The relevant 10 legislative framework is set out at [11] to [19] above. While subsections (2) and (3) of section 84A exclude certain exchange gains and losses from the loan relationship regime, the effect of Regulation 13 of the 2002 Regulations is to reverse that exclusion in the circumstances set out in that regulation. Under Regulation 13, the aggregate of the exchange gains and losses determines the company’s “net gain or net loss”, which 15 is then brought into account for corporation tax purposes as a loan relationship credit (for a net gain) or debit (for a net loss). Exchange gains and losses are defined by section 103(1A) as follows: (1A) References in this Chapter to exchange gains or exchange losses, in the case of any company, are references respectively to— 20(a) profits or gains, or

(b) losses, which arise as a result of comparing at different times the expression in one currency of the whole or some part of the valuation put by the company in another currency on an asset or liability of the company. 25 If the result of such a comparison is that neither an exchange gain nor an exchange loss arises, then for the purposes of this Chapter an exchange gain of nil shall be taken to arise in the case of that comparison. The wording of the definition is clear and unambiguous. It does not refer in the abstract to “profits” or “losses” in relation to foreign currency. Nor does it refer, 30 explicitly or by implication, to profits or losses measured by reference to some other criterion such as exposure. It refers specifically to gains or losses “which arise as a result of comparing at different times” the relevant valuations. It is plain that the definition requires, and in our opinion only requires, a comparison to be made at two different times. Whether the product of that comparison 35 is or is not a “purely arithmetical difference” is irrelevant to this question. If the comparison produces a loss (or gain), then it is an exchange loss (or gain), because it “arises as a result of” the comparison mandated by the statute. The suggestion that the draftsman intended to incorporate additional unspecified criteria into the definition is precluded both by the statutory framework and the unambiguous wording of the 40 legislation. 26 The second part of subsection (1A) demonstrates that it is the comparison which determines the existence for corporation tax purposes of an exchange gain or loss within section 84A and a net gain or net loss within the 2002 Regulations. Neither the taxpayer nor HMRC has discretion to claim that exchange gains or losses arise outside the code. If the comparison shows that in aggregate neither an exchange gain nor 5 exchange loss arose, then the “exchange gain of nil” taken to arise is subject to the exclusivity of section 80(5). The FTT was right not to be bound by the decisions in Stagecoach and Union Castle, because they concerned the loan relationships code and derivative contracts code respectively. An exchange gain or loss arises solely as a result of the comparison 10 required by section 103(1A), subject to the safeguards found in the need for GAAP compliance and the “fairly represents” requirement. It requires no action on the part of the company or another person, and, unlike a loan relationship or derivative contract, does not relate to a bilateral agreement. Even leaving these differences aside, the decision of this Tribunal (of which Fancourt J was a member) in The Union Castle 15 Mail Steamship Company and the Commissioners for Revenue & Customs

[2018] UKUT 316 (TCC)

illustrates the importance of construing the statutory language in its specific context. The Upper Tribunal decision, delivered after the hearing of this appeal, reversed the finding of the FTT that there was no “loss” on the basis of a detailed contextual analysis of the particular code. 20 Two further points reinforce our interpretation of exchange loss. First, the comparison applies equally to gains and losses, and in our opinion it is a sensible outcome that a taxpayer for whom the comparison produces a gain, in a situation where the relevant accounts are GAAP compliant and it is accepted that the exchange difference “fairly represents” a gain, should not be able to remove the profit from 25 corporation tax on the basis that it is in some way not a “real” or “ actual” profit, or accurately reflective of exposure. Secondly, the safeguards which HMRC seek in relation to this issue are provided by the statutory code and in the GAAP compliance and “fairly represents” requirements. In view of our decision that HMRC’s appeal on the second issue is dismissed for 30 the reasons given, it is unnecessary to consider at this stage the extent to which “real” or “economic” consequences arose, but we do so below in relation to the third issue. Third issue: “fairly represents” HMRC appeal against the FTT’s decision that the exchange losses did “fairly represent” losses of the Companies. 35 The issue The relevant accounts of the S&N Companies were prepared in accordance with GAAP, and the relevant losses were “exchange losses”. However, section 84(1) (set out at [14] above) further stipulates that the credits and debits to be brought into account in respect of a company’s loan relationships shall be “the sums which, when 40 taken together, fairly represent, for the accounting period in question…all profits, gains and losses of the company…which arise to the company from its loan relationships and related transactions”. 27 HMRC argue that the FTT was wrong to conclude that the losses in this case did “fairly represent” losses of the company, because the companies had no underlying foreign exchange exposure and suffered no real economic loss. Discussion As set out at [8] above, at the time of hearing this appeal the Court of Appeal had 5 not yet handed down its judgment in GDF Suez. That decision was released on 5 October 2018 and is reported at

[2018] EWCA Civ 2075

. Our discussion of this issue takes account of the submissions which we allowed the parties to make, and which they did make, following the release of the decision. The meaning and effect of the words “fairly represent” have been considered in 10 several cases. None of the decisions concerns the phrase as used in the context of the foreign exchange provisions in issue in this appeal, though section 84A(1) assimilates them to the loan relationships code in the Finance Act 1996, as amended. In Commissioners for Revenue & Customs v DCC Holdings (UK) Limited

[2010] UKSC 58

, the Supreme Court considered the test in relation to a purchase and sale (or 15 repo) of gilts. In its decision to restrict the deduction claimed by the taxpayer, the Supreme Court referred to the absurdity of the asymmetrical result which would otherwise arise, and emphasised that “the need for a symmetrical solution lies at the heart of this appeal”: see paragraph [26] of the judgment. In Greene King plc and anor v Commissioners for Revenue &Customs [2016] 20 EWCA Civ 782, the Court of Appeal dealt with the test as one of several issues in relation to a marketed tax avoidance scheme, which sought to exploit the loan relationship rules. The taxpayer in that case sought to avoid a charge to corporation tax on a loan relationship credit on the basis that it did not fairly represent a profit because “it was not a profit or gain in any meaningful sense”: [35] of the decision. 25 The Court, referring to DCC Holdings, rejected the taxpayer’s argument, partly due to lack of evidence. We turn now to GDF Suez. The judgment is, of course, binding on us for what it decides. In its analysis of the “fairly represents” issue, the Court considered it essential to note at the outset that the version of section 84(1) considered in DCC Holdings and 30 Greene King was that in force before the amendments made by the Finance Acts 2004 and 2006: see paragraph [74] of the decision. Given the Court’s conclusion that the legislative history was of material importance in interpreting the provision, we should be cautious in simply reading across what was said on the issue in DCC Holdings and Greene King, given that this appeal concerns the same (amended) version of section 35 84(1) as in GDF Suez. We consider that three propositions of general application can be drawn from GDF Suez in interpreting the “fairly represents” requirement of section 84(1) as it stood following 2006. They are as follows: (1) First, particularly in view of the legislative history, it is intended to 40 operate as an override. In 2006 the GAAP requirement of section 85A was amended so that it came to read “[S]ubject to the provisions of this Chapter 28 (including, in particular, section 84(1))…”. The Court concluded, having taken into account the Government’s Explanatory Notes to the amendment, that the purpose of this change was to address the mischief of tax avoidance schemes. It did this by “making it absolutely clear that the “fairly represents” rule in section 84(1) takes priority over, and may override, the accounting 5 treatment mandated by section 85A(1)”: [50] of the decision. It therefore imposes a separate and additional requirement for recognition of profits and losses: [88] to [91] of the decision. (2) Secondly, section 84(1) is not limited in its purpose or effect to 10 attribution or allocation of gains and losses to accounting periods or to loan relationships and related transactions: see [92] of the decision. (3) Thirdly, the absence of specific statutory guidance as to the test need not be problematic. In responding to a submission by Mr Ghosh (who also represented the taxpayer in GDF Suez) the Court stated: 15 “93. The objection that Parliament would have formulated specific guidance on the application of the fair representation test, if it was intended to be an overriding requirement of a substantive nature, is at first sight more compelling, particularly when it is remembered that the test was until 2004 explicitly linked to “an authorised accounting 20 method”. Nevertheless, I do not think that the objection is well-founded, although it was persuasively advanced by Mr Ghosh. The concept of fairness is central both to the development and application of accounting standards, and to any process of judicial appraisal by a court or tribunal. In itself, the concept needs no elucidation, but rather provides a 25 touchstone which is well suited to application by accountants, lawyers and judges, bringing their professional experience and expertise to bear in widely differing factual contexts”. We turn now to the FTT’s decision on this issue. It is set out in the Decision as follows: 30 “50. We therefore agree with the [FTT] in Union Castle…that the words “fairly represent” have been included in the legislation for an identification and/or timing purpose to identify from entries in the accounts those credits or debits which arise, in that case from derivatives and [in] the present case a company’s loan relationships, and which 35 entries are appropriate in a particular accounting period.

[51]Accordingly, we find that the Exchange Differences do “fairly represent” a loss arising to the appellants, as defined by s 84(1), during the period ending 31 December 2008.” Regardless of the Court of Appeal’s decision in GDF Suez, we would not have 40 found the FTT’s reasoning to be correct. This Tribunal had rejected the limited attribution/timing proposition in Union Castle, at paragraph [47] of that decision, before the release of the GDF Suez decision. As the second proposition set out above makes clear, section 84(1) is not so restricted. The FTT therefore erred in law in basing its conclusion on this reasoning. That 45 does not, however, mean that we must set aside the Decision on this issue. Section 12 29 TCEA states that we may set it aside. We need not set it aside if we conclude that, notwithstanding the error of law in its reasoning, the FTT reached the correct conclusion on the issue. In their submissions following the release of GDF Suez, HMRC argued as follows: 5(1) The decision means that the test of fair representation requires an assessment of whether a loss arose as a matter of fact or general principle.

(2) Applying that approach, there was no real loss because there was no economic exposure to foreign exchange movements during 2007 or 2008.

(3) Similarly, there was no real loss because the exchange differences were 10 not “realised” losses but merely “presentational losses arising on re-calibration”. This was reflected in the treatment of distributable reserves in relation to the dividend paid by S&N Overseas in December 2008.

(4) There was a mismatch because the exchange differences were not recognised in the S&N group’s consolidated accounts, providing another 15 reason why the differences did not fairly represent losses.

(5) The wider context suggested there was no real loss. In particular, the transactions were entirely intra-group, and the share premium issued as part consideration was subsequently reduced by a capital reorganisation.

(6) The section 84(1) override “should operate to address anomalies that 20 arise as a result of the accountancy treatment and/or the mechanical operation of the statutory regime”. Counsel for the S&N Companies submitted that GDF Suez did not have the effect put forward by HMRC, for the following reasons: (1) The appreciation of the US dollar against sterling had a real commercial 25 impact on the companies, irrespective of the accounting treatment, in the amount of the additional share premium payable by S&N Overseas. (2) The adoption of the FO Method reflected economic and commercial reality. (3) The recognition of exchange losses followed from the fact that 30 Parliament had mandated that a foreign exchange loss must be measured by reference to valuations at two dates. It is irrelevant whether the loss has been realised. Further, Regulation 13(3) specifically requires that an amount equal to the net loss be brought into account as a loan relationship debit. (4) GDF Suez seemingly restricts the role of the “fairly represents” 35 mechanism to tax avoidance cases. This is not such a case. In considering the third issue in this appeal, we draw guidance from GDF Suez in two respects. First, we bear in mind the three general propositions set out at [81] above. Secondly, while GDF Suez turns on its facts, it does serve to give a clear indication of the most significant factors to which a court is likely to give weight when 40 considering all the circumstances and determining whether to apply the override. We 30 identify the following factors as significant in the decision by the Court of Appeal to apply the override in order to bring the credit on disposal in that case into account: (1) The case concerned a tax avoidance scheme. (2) The effectiveness of that scheme was, if not counteracted by the override, clearly contrary to Parliamentary intention: see [96] and [99] of 5 the decision. (3) Taxing the credit avoided asymmetry for tax and accounting purposes between transferor and transferee: [96]. (4) Without an override, the claims would fall out of charge to tax altogether: [96]. 10 (5) The transferor could fairly be regarded as having made a profit or gain on the disposal: [97]. (6) On the facts, the requirement to apply the override was obvious (“the proof of the pudding is in the eating of it”): [95]. The facts of this appeal are very different from those in GDF Suez. First, this case 15 does not concern a tax avoidance scheme. Indeed, no tax avoidance motive on the part of the S&N Companies was alleged or found on the facts. While we reject Mr Ghosh’s proposition that GDF Suez can be taken as authority that the override can only ever apply in a tax avoidance situation, we do consider that the presence of a tax avoidance motive on the part of a taxpayer is likely in practice to be the paradigm situation in 20 which it may be justified (depending on all the facts and circumstances) to apply the section 84 override. Such an approach is in our view borne out by the weight placed by the Court of Appeal on the purpose of the 2006 amendment to section 85A, which was to counter the mischief of tax avoidance arrangements which sought to rely on a narrow interpretation of the override. 25 Secondly, there is in this case no accounting mismatch of the kind which arose in GDF Suez and DCC Holdings. HMRC’s argument that a mismatch arose because the loss from the exchange differences was not reflected in the consolidated accounts of the group is misconceived. The group’s consolidated accounts were prepared in US dollars before the group reorganisation which led to the change in functional currency 30 of the S&N Companies, so need not have recorded the change to US dollars within a sub-group, and given that the relevant receivables were intra-group the foreign exchange exposure would in principle net out on consolidation in any event. At a broader level, the point is misconceived because an exchange gain or loss, being a product of a comparison at two different times, will not necessarily generate a 35 “symmetrical” loss or gain in the same way as (for example) a disposal of a loan relationship. Thirdly, the application of the override in this case is not required in order to avoid a result which is clearly contrary to Parliamentary intention. As described above, the regime for exchange differences mandates, in section 103(1A), the method 40 by which foreign exchange gains and losses must be calculated. Regulation 13 then specifies that an amount equal to the net gain or loss “shall” be brought into account as a credit or debit. While we do not accept Mr Ghosh’s argument that these provisions effectively preclude the application of the override, the provisions do not support the 31 proposition that the recognition of the losses in this case, in accordance with GAAP, is (as in DCC Holdings and GDF Suez) clearly contrary to the intention of Parliament. Fourthly, the override is not a mechanism to re-open the status of the accounting method for GAAP purposes. The result produced in this case by the GAAP compliant accounts was not, in the language of DCC Holdings, a manifest absurdity, such that 5 the need to apply the override was obvious. HMRC raised a number of arguments to support the application of the override based on the commercial or company law consequences of the change in functional currency. We refer to them above. We have concluded that those arguments either broaden the remit of the override to an extent which is unwarranted or consist in 10 substance of a renewed attack on the FO Method. The existence of differences between the calculation for tax purposes of an exchange gain or loss and the treatment for company law purposes of that gain or loss in the calculation of distributable reserves is not necessarily an indication that the tax calculation is not a fair representation of the loan relationship gain or loss. The 15 presence or absence of exchange exposure is, as we have explained above, not a benchmark which the comparative calculation requires or incorporates. Indeed, HMRC’s arguments on this issue come very close to a submission that the FO Method can never result in a fair representation. The additional share premium issued by S&N Overseas as part consideration illustrates that the change in functional currency did 20 have “real world” consequences. The fact that the premium may have been partially reduced five years later is nothing to the point. Finally, the fact that the transactions were intra-group is not in our opinion a reason in favour of applying the override. Gains and losses can arise within a group, and the legislation contains many detailed provisions relating to group transactions. 25 Taking into account all the facts, including but not limited to the absence of a tax avoidance motive, the absence of any material asymmetry, and the absence of an absurd result, we have concluded that the exchange losses in this case did fairly represent losses as required by the legislation. Disposition 30 For the reasons given, the appeal in relation to the first two issues is dismissed. In relation to the third issue, we affirm the decision of the FTT but for different reasons and accordingly the appeal on the third issue is also dismissed. MR JUSTICE FANCOURT 35 JUDGE THOMAS SCOTT RELEASE DATE: 29 November 2018 40

The Commissioners for HM Revenue and Customs v Smith and Nephew Overseas Ltd and Others: [2018] UKUT 0393 (TCC) [2018] UKUT 0393 (TCC)

Cited in 2 later judgments