Ardmore Construction Ltd v The Commissioners for HMRC [2018] EWCA Civ 1438

[2018] EWCA Civ 1438Case No A3/2016/0688
IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM
UPPER TRIBUNAL (TAX AND CHANCERY CHAMBER)
MR JUSTICE MORGAN & JUDGE ROGER BERNER
[2015] UKUT 633 (TCC)
Venue Royal Courts of Justice, Strand, London, WC2A 2LLDate 21/06/2018
Before
LADY JUSTICE ARDEN LORD JUSTICE SALESLORD JUSTICE LEGGATT
Between
Ardmore Construction LimitedAppellantThe Commissioners for Her Majesty’s Revenue and CustomsRespondent
Patrick Way QC (instructed by Sharpe Pritchard LLP) for Appellant Richard Vallat QC (instructed by HMRC’s Solicitor’s Office) for the RespondentHearing Hearing dates : 14 March 2018
Judgment ApprovedLADY JUSTICE ARDEN :
[1]This is an appeal by Ardmore Construction Limited (“Ardmore”) about the correct application of the “source” principle, that is the principle that tax is territorial and income tax, for instance, is only payable on income from a source in the UK (see now section 368 of the Income Tax (Trading and Other Income) Act 2005 (“ITTOIA”), set out in paragraph 3 below). The principle is long-standing: see in Colquhoun v Brooks (1889) 14 App Cas 493, 504. The issue is how the source principle is to be applied to interest paid on a foreign loan in this case. There is no universal test for applying the source principle: in this case, using the language of the day, the test has been described as “multifactorial” and thus as involving an overall assessment of the situation.[2]This appeal is from the determination of the Upper Tribunal (Morgan J and Judge Berner) dated 20 November 2015. Mr Patrick Way QC appears for Ardmore and Mr Richard Vallat QC appears for HM Revenue and Customs (referred to below as “HMRC”). This judgment will set out some unusually long citations from three decided cases to see how courts have applied the principle in different circumstances. These citations are instructive in showing how the courts have attributed weight to different factors.

Legislative framework

[3]The material provisions of section 874 of the Income Tax Act 2007 (“ITA”) and section 368 of ITTOIA are as follows: 874 Duty to deduct from certain payments of yearly interest(1) This section applies if a payment of yearly interest arising in the United Kingdom is made— (a) by a company,… (b)(2) The person by or through whom the payment is made must, on making the payment, deduct from it a sum representing income tax on it at the basic rate in force for the tax year in which it is made…. 368 Territorial scope of Part 4 charges (1) Income arising to a UK resident is chargeable to tax under this Part whether or not it is from a source in the United Kingdom. (2) Income arising to a non-UK resident is chargeable to tax under this Part only if it is from a source in the United Kingdom. (2A) …(3) References in this section to income which is from a source in the United Kingdom include, in the case of any income which does not have a source, references to income which has a comparable connection to the United Kingdom.(4) …

Background

[4]The facts are agreed and are fully set out by the Upper Tribunal at paragraphs 69 to 80 of its judgment. A summary suffices here. Ardmore is a UK resident company owned and managed by two brothers. Its assets and activities are almost entirely in the UK. Using working capital, it subscribed for shares in two companies incorporated in the British Virgin Islands (“BVI”) and owned by Gibraltar trusts established by the brothers respectively. The detail of the trusts and the terms of the shares subscribed is immaterial. The sum of £1.35m, being a sum equal to the amount subscribed (by way of part payment of share premium) by Ardmore to the BVI companies, was lent by those companies to the trusts and by the trusts to Ardmore, in the latter case under a facility letter dated June 2005.[5]The terms as to interest and repayment of capital were unremarkable. The relatively unusual features of the loan were that(1) it was not required not to be supported by any UK security,(2) the facility letter and loan documentation were governed by the law of Gibraltar and the parties submitted to the exclusive jurisdiction of the Gibraltar courts, and(3) (although this did not happen) payments of interest were to be made in Gibraltar from a non-UK source.[6]HMRC determined that the interest paid on the loan was paid from a source within the UK and that Ardmore ought to have deducted tax from it. Ardmore appealed to the First-tier Tribunal (“FTT”). The Upper Tribunal heard the appeal in this case with that in Perrin v HMRC, another decision of the FTT, and the judgment of the Upper Tribunal relates to both cases. However, the latter appeal is not before this Court and I need not refer further to it.

Decisions of the First-tier and Upper Tribunals

[7]Ardmore submitted that the appropriate test was that of the place where the lender was. By their decision dated 21 May 2014, the FTT (Judge Brooks and Mr Stafford) rejected Ardmore’s appeal. They considered several cases on the source principle. For example, they applied the speech of Lord Nolan, giving the advice of the Privy Council in Commissioner of Inland Revenue v Orion Caribbean Limited (in voluntary liquidation) [1997] STC 923, where he found that the proposition enunciated by Lord Bridge in Commissioners of Inland Revenue v Hang Seng Bank Ltd [1991] 1 AC 306 PC to the effect that the source of income was always located in the place where money is lent could not stand with the line of authority establishing that the ascertaining of a source of income is a “practical hard matter of fact” (to use the words of Lord Atkin in Rhodesia Metals Ltd (in liquidation) v Commissioner of Taxes [1940] AC 774, 789). Lord Nolan held that: No simple, single, legal test can be employed.[8]Ardmore appealed to the Upper Tribunal. Ardmore submitted that(1) the source of the interest could properly be found by ascertaining the nationality or residence of the relevant loan instrument,(2) if the proper test was the multifactorial test, the FTT erred by affording too much weight to the residence of the debtor, and(3) the source of the interest was the place where the credit was provided.[9]The starting point for Ardmore’s submissions was that the underlying tax liability was that of the lender, and not that of the borrower. Ardmore submitted that the proper approach was to step back and ask whether the loan document was fundamentally a UK document or a foreign document.[10]The Upper Tribunal also considered several authorities, particularly Westminster Bank Executor and Trustee Co (Channel Islands) Limited v National Bank of Greece SA (1970) 46 TC 472 (“the National Bank case”), which was the only binding authority on the meaning of the statutory phrase “interest arising in the” UK. In that case, the source principle had to be applied to the following critical facts. Bearer bonds subject to English law, denominated in sterling and secured (on issue) on property in Greece were (following amalgamation and universal succession under Greek law) the subject of an unconditional guarantee by the National Bank (which was non-resident). The coupons fell to be paid by the guarantor through paying agents in London or at the option of the bearer of the coupon in Greece by cheque drawn on London. The funds to pay the interest would have to be remitted from Greece to London.[11]Lord Hailsham, with whom the other members of the House of Lords agreed, approached the matter experientially by weighing and comparing the factors, and this process has become known as the multifactorial test: I have come to the conclusion that the source of the obligation in question was situated outside the United Kingdom. This obligation was undertaken by a principal debtor which was a foreign corporation. That obligation was guaranteed by another foreign corporation which, as was conceded before us, had at no time any place of business within the United Kingdom. It was secured by lands and public revenues in Greece. Payment by the principal debtor of principal or interest to residents outside Greece was to be made in sterling and either at the offices of Hambros Bank or Erlangers Ltd or (at the option of the holder) at the National Bank of Greece in Athens, Greece, by cheque on London. Whichever method of payment was selected it was pointed out before us that, whatever use were made of the option, discharge of the principal debtor's obligation would have involved in the ordinary course either a remittance from Greece to the paying agents specified in the bond or, at the option of the holder, a cheque issued within Greece although drawn on London and presumably payable there out of funds remitted by the debtors from abroad. It was also pointed out that the bond contained no provision for payment by the guarantor at any particular place or in any particular country. The only circumstances relied on by the appellants as supporting their contention that the obligation was located inside the United Kingdom were as follows. Although the original guarantor had no branch in the United Kingdom the appellants had acquired one, on the universal succession in London. Moreover it was urged that since discharge of the obligations under the bond in Greece had been caught by the moratorium enacted by the Greek government, it followed that the only place at which the obligation could have been discharged or enforced was in London. Speaking for myself, I do not see how an obligation originally situated in Greece for the purposes of British income tax could change its location either by reason of the fact that one guarantor had been substituted for another, or by the fact that the second guarantor so substituted subsequently acquired a London place of business or by the fact that the government of Greece had by retrospective legislation altered by moratorium and substitution of a new guarantor for the purposes of Greek law the obligations imposed on the principal debtor and the guarantor. The appellants acquired no obligation different from that of the original guarantors, and that was the obligation imposed on the original guarantors by the terms of the bonds. In my view the bond itself is a foreign document, and the obligations to pay principal and interest to which the bond gives rise were obligations whose source is to be found in this document.[12]Lord Hailsham added the following passage (not cited by the Upper Tribunal) which shows the importance of the connection with Greece as a matter of fact in terms of the residence of the debtor and original guarantor and source: We were much pressed in this House by the appellants with the case of Foulsham v. Pickles [1925] A.C. 458 which, it was contended, was an authority for the proposition that the source of income in such a case as this is situated in the United Kingdom. Pickles v. Foulsham was not cited to either of the courts below and so we have not the advantage of the judgment in those courts. It is in fact a case in which some of the observations there made in this House may one day be said to have gone too far in one direction or another. But in any event it is a case which stands on its own facts which are vastly different from the present. In that case a contract was made in the United Kingdom to pay money in the United Kingdom by a British company for services to be rendered in West Africa by a resident of the United Kingdom. Income arising under this contract was held not to be income the source of which was wholly situated abroad. In the instant case both the principal debtor and the original guarantor were wholly resident abroad, and the source from which the income was contemplated to be paid was certainly intended to be remitted from abroad and was partly secured by property situated abroad. On the whole I agree with Finlay J. in Chamney v. Lewis (1932) 17 T.C. 318, 323 that Foulsham v. Pickles is more an authority for the wide meaning of the term "possession" for the purposes of Case V than a reliable guide to the situation of a source of income on facts which differ from those then in question.[13]The Upper Tribunal rejected Ardmore’s argument (1) (paragraph 8 above) because it focused on a single criterion, and this was inconsistent with the multifactorial approach in the National Bank case. The place where the credit was provided could not be treated as the single relevant factor.[14]The Upper Tribunal permitted Ardmore to withdraw a concession that, if the test was multifactorial, it could not succeed. Nonetheless, it held that applying that test the source was in the UK: [90] We do not accept those submissions [i.e. arguments (2) and (3) above]. They do not accord with the law as authoritatively established in the Greek Bank case, as we have described. The FTT was right to give weight to the residence of the debtor, Ardmore, which was in the UK. It was right to have regard to the substantive, and in this case actual, source of the payments, which derived from Ardmore's UK trading activities. For the reasons we have described, we agree with Mr Vallat that the residence of the lender and the place from which the money was lent (the place of credit) are not relevant. Jurisdiction and the proper law of the contract are factors of little or no weight having regard to [ the National Bank case]. The absence of security is a neutral factor; the fact that the agreements specified that there should be no UK security is immaterial.

Submissions and discussion

[15]Matters have moved on: Ardmore now accepts that under the source principle the test is multifactorial but it maintains that in this case the source is in the creditor’s place of residence to which greater weight should have been given.[16]There are two candidates for the source in this case: Gibraltar and the UK. The connecting factors in the case of Gibraltar are the place where the loan was made, the place of the trusts’ bank accounts, the governing law and exclusive jurisdiction clauses, the place of residence of the creditors and the absence of security in the UK. The connecting factors with the UK are the physical source of the funds for paying the interest, the location of the assets on which any judgment given in Gibraltar would be enforced, the business of the borrower and Ardmore’s bank accounts and thus the place from which interest was paid, and the absence of any security over foreign property. The parties are agreed that the situs of the debt (that is, the place where a debt is located) for the purpose of the conflict of law rules is of little or no weight as the rules as to situs are legal rules only.[17]Ardmore argues that Gibraltar is the source of the interest. Mr Way submits that the Upper Tribunal was wrong to consider that the single most important factor was the debtor’s residence. They should have found that Gibraltar was the source because that was the place of the creditor’s residence, the place from which credit was given, as the place of exclusive jurisdiction and the place of the creditor’s activity. The debtor’s place of residence is relevant only when it is the place where the creditor would sue him. The withholding of tax by the borrower is on account of the lender’s activities and so the focus should be on the lender. Absent the provision of credit, there would be no interest. The lender had no place of business in the UK. The exclusive jurisdiction clause made the debtor’s UK residence irrelevant.[18]Mr Way also submits that the weight which the Upper Tribunal gave to the residence of the debtor makes for difficulty where the debtor is dual resident because it would mean that there was no source. Moreover, the debtor’s residence is an inappropriate factor especially where there is an exclusive jurisdiction clause. In addition, no security has been given in the UK.[19]Mr Way submits that the New Zealand case of Commissioner of Inland Revenue v NV Philips Gloeilampenfabrieken [1955] NZLR 868 (NZ CA) (“the Philips case”), considered by the Upper Tribunal, though only persuasive authority, supports his argument. The issue in the Philips case was whether interest paid by a New Zealand company to a lender in the Netherlands pursuant to a loan raised in the Netherlands and governed by Dutch law was the Netherlands for the purpose of section 87(n) of the Land and Income Tax Act 1923 of New Zealand. Section 87(n) rendered assessable to income tax “income derived directly or indirectly” from money lent in New Zealand. The New Zealand Court of Appeal held that the source did not mean the place from which monies came, but the originating cause of the payment being made. Gresson J, delivering the first judgment of the Court of Appeal, gave an illuminating judgment about determining source (applying the practical approach described below). He held, relying on the judgment of Watermeyer CJ in the South African case of Commissioner for Inland Revenue v Lever Brothers, that the “source” was the originating cause of the interest paid on a loan was the place where the loan was made: The ordinary meaning of “source” is the starting-point which, when used in relation to physical things, e.g., a river, is a matter of location. But it is a word of flexible meaning, especially when used of something non-material or abstract. It can, and often does, mean the chief or prime cause of something. What has to be determined is the sense in which the Legislature used the word in s. 87 (n). The test-what a practical man would regard as the real source as a practical, hard matter of fact-which was formulated in Nathan v. Federal Commissioner of Taxation ((1918) 25 C.L.R. 183), approved as it has been by the Privy Council in Liquidator, Rhodesia. Metals, Ltd. (In Liqdn.) v. Commissioner of Taxes ([1940) A.O. 774; [1940) 3 All E.R. 422), must be adopted. The answer which I should expect the " practical man '' to make to a question - What was the source of the money which was received by the Dutch company?-would be the loan it made which means, in effect, the lending of the money-the transaction. The money was paid because the New Zealand company had contracted to pay it; so that, in some sense, it can be said the obligation which had been entered into was the source of the payment made. But one must look behind that. It is seldom that a person makes a payment except under an obligation to do so, and it is, I think, unreal and incompatible with a practical approach to regard the obligation as the source. It is what produced the obligation that is important. A lessee pays rent because he has entered into an obligation to do so, but he has only done this on terms that land is made available to him. An o b l i g a t i o n is seldom, if ever, accepted in v a c u o : it requires some transaction to give it birth. The obligation arises from something which has been, or will be, done to warrant it, e.q., rendering services, making land or other property available. The practical man, in regarding the loan as the source of the payment, would mean, I think, the conduct or the action which was the reason for the obligation being accepted. The document executed stated that the loan had been made, and that was the originating cause of the payment of interest. That was the view taken by Watermeyer, C.J., in the South African case of Commissioner for Inland Revenue v. Lever Brothers and Unilever, Ltd. (1946) 14 S. Af. Tax Cas. 1) – that “source” does not mean the quarter whence the moneys come, but the originating cause of the payment being made – the quid pro quo which the recipient of the money gave to entitle him to receive payments from time to time; that in the case of a loan, the lender provides money for the borrower, who, in return, pays interest until such time as he makes repayment:
“The lender either gives credit to the borrower or transfers to him certain rights of obtaining credit which had previously belonged to the lender, and this supply of credit is the service which the lender performs for the borrower, in return for which the borrower pays him interest. Consequently this provision of credit is the originating cause or source of the interest received by the lender. Although, colloquially, one speaks of a debt carrying interest, or interest on a debt, as though interest were a sort of growth sprouting from the debt, the language used means no more than that the borrower pays interest, if that is the agreement between borrower and lender, as consideration for the benefits allowed to him by the lender” (ibid., 10). Davis, A.J.A., in agreement with Watermeyer. C.J., said: “The practical man would say that the source of Levers’ income was the provision by it of assets in America and the giving of credit in England” (ibid., 23). Schreiner J.A., disagreed, his view being that when income is derived by a person from another who is using that person’s property, and that property happens to be money, the interest is derived from the loan, the local situation of which must be ascertained. I think the decision of the majority is to be preferred. It appears to me that in interpreting s. 87 (n), proper regard must be paid to the word “derived”; it should not be read as “received”
. The word “derived” means more than received; it connotes the source or origin, rather than the fund or place, from which the income was taken. It means flowing, springing, emanating from, or, as was said in Commissioners of Taxation v. Kirk ([1900] A.C. 588, 592), arising from or accruing. To be a “source” of the income within the meaning of the subsection, it is necessary, I think, to look to the originating cause. It is not sufficient to ascertain the fund out of which the income was in fact paid, which is no more than the reservoir from which it was drawn. It is not whence it was paid, but why it was paid, that is the determining factor. The emphasis is not upon the receipt, but upon the derivation of the income. Consequently, it does not constitute the source within the meaning of the section that the money was drawn from or provided by the trading profits in New Zealand. The New Zealand company was free to obtain the funds with which to perform its obligation anywhere it chose, from deposits in England, if it had any, or from borrowing in England, or from the profits of its trading in New Zealand. That was a domestic matter. The money could "come from" any of these “sources”, but none of them would be the source from which the Dutch company derived what it received as income. The combination of the words “derived” and “source” import, I think, some causative link. In my view, therefore, the originating cause being that the Dutch company had lent moneys or provided a credit in London, from which sprang the obligation to pay interest, the “source” of the Dutch company’s income, was not in New Zealand, even though the borrower resorted to its New Zealand funds to pay the interest. Where it got the money with which it in fact paid the interest is, I think, irrelevant. In a physical sense the money came from the trading activities in New Zealand; but that was a domestic matter. Looking at the real substance of the facts with the eyes of a practical man, it was from the provision of the loan moneys that the income was derived; the title to the money paid as income sprang from the loan. The money, in fact, came from New Zealand. But the statute does not say “received from a source in New Zealand”, but “derived… from [a] source in New Zealand”: s. 87 (n). In my opinion, the appeal fails. (paragraph breaks added)[20]Mr Way submits that his case is also supported by the fact that in relation to capital gains tax there is a special provision in section 275(1) of the Taxation of Capital Gains Act 1992 (“the TCGA 1992”) which provides that, for the purposes of that Act, subject to provisions of section 275(1), the situs of a debt (whether secured or unsecured) is in the UK if (and only if) the creditor is resident in the UK.[21]Mr Way also submits that the multifactorial test applied by Lord Hailsham should be applied in light of modern circumstances. In contrast to the National Bank case, a court should ignore the residence of the debtor where there is an exclusive jurisdiction clause, and should take into account the exclusive jurisdiction as a factor in itself, as well as the residence of the creditor is highly relevant.[22]Mr Vallat submits that ascertaining the source of income for tax purposes is not as simple as following the money used to pay the interest. The correct approach was to have regard to the “underlying commercial reality” and seek the substantive source of payment. In this case all the real substantive connecting factors connected the interest with the UK. The activities which generated the monies to pay the interest were in the UK. If merely non-UK income had been used to fund the interest, there might have been a different conclusion. The relevant test is that proposed by Mr Warner (as he then was) in the National Bank case (the test called in these proceedings the multifactorial test). It involves consideration of a number of factors including the origin of the monies, the residence of the debtor, the difficulties of enforcement, security, place of residence of guarantor, etc.[23]Moreover, submits Mr Vallat, an appellate tribunal should not interfere with a multifactorial assessment unless it discloses an error of law. There is no error of law in the approach adopted by the Upper Tribunal in this case; it properly directed itself in law and considered all the relevant matters. Ardmore is wrong, submits Mr Vallat, to suggest that the Upper Tribunal reduced the question to a simple test depending on residence. The Upper Tribunal was right not to attach weight to the residence of the lender (and it follows, the place of credit), the fact that there were two offshore companies and the situs of the debt, and to conclude that the interest arose in the UK because the residence of the debtor is an important, but not paramount, factor. The residence of the lender is irrelevant, as is the fact that there were two offshore trusts and two offshore companies. There is no evidence of any substantive business activity by the companies. In any event, as simple contract debt, it is located where the debtor is located; the UK. The proper law of the obligation can be given very little, if any, weight. The place of exclusive jurisdiction may be relevant, but its importance is reduced if the place of substantial enforcement is different. Had Ardmore defaulted, the place of ultimate enforcement would have been the UK. The only funds available to Ardmore to pay the interest were in the UK. The place of actual or stipulated payment may be relevant, but of little weight.[24]Mr Vallat submits that, leaving aside authority, the residence of debtor and guarantor, and substantive source of payment and location of security are appropriate factors as they are based on the underlying commercial reality of the interest payments rather than legal factors capable of change and manipulation. Factors of little or no weight include the residence of the creditor (which he submits was not mentioned by Lord Hailsham, but this is not correct: see paragraph 12 above); the location of the bank accounts from and to which the funds were advanced (mentioned by Lord Hailsham); the place of payment of the interest; the jurisdiction in which proceedings might be brought to enforce the interest; the proper law of the contract and the situs of the debt.[25]Ardmore is also, on Mr Vallat’s submission, wrong to contend that the only relevant factor is the place where the creditor could sue to recover the loan or interest on it. In the National Bank case, the House of Lords had already determined that legal proceedings for the National Bank case could only be brought before the courts in England and Wales. If jurisdiction is determinative, the National Bank case would be wrongly decided. The true position is that the House was concerned with substantive factors based on the underlying commercial reality.[26]Mr Vallat submits that it is wrong to characterise the tax charge as arising due to the lender’s activities. He submits that, tax is due because interest arising in the UK is paid by the debtor; there is no need for any creditor activity at all. The source of the interest and therefore the reason for the obligation to deduct tax is the loan, not the activity of the creditor. The lender is purely passive.[27]The judgment of Lord Hailsham is inconsistent with the causative test used in Phillips. The whole background in the National Bank case was the debt could only be enforced in the UK courts. Residence was not relevant. All the factors pointed to Greece and the factors pointing to the UK were insignificant. The transaction was “innately foreign”. It was a Greek business and a Greek moratorium. The fact that change had occurred since the bonds were issued was irrelevant but the point of principle established in National Bank goes wider than that. In this case the residence of the debtor, the substantive origin of the funds all point to the UK as the source. There was no foreign security (c.f. the National Bank case).[28]There is nothing to support the place of credit as the test. The Commonwealth cases must be treated carefully because they are decided under different legislation. It is necessary to look at the matter substantively to prevent manipulation. That is one reason why Ardmore’s test cannot be accepted.[29]Moreover, submits Mr Vallat, the creditor could sue Ardmore on the loan in the UK because it was resident here and this rule is not affected by the Brussels I Regulation or the Lugano Convention: see Dicey, Morris and Collins, The Conflict of Laws, (15 ed, 2017) volume 2, Rule 129 at 22RE-023 onwards, especially 22-026 and 032. Residence should thus be treated as a significant connecting factor for the purposes of ascertaining where interest arises even if there is an exclusive jurisdiction clause. The Upper Tribunal therefore reached the correct conclusion that residence was a relevant factor regardless of restrictions on jurisdiction.[30]Dual residence, on Mr Vallat’s submission, would not cause any great difficulty in the application of the multifactorial test, as the weight given would be affected by the basis of dual residence status. This is consistent with assessment based on a commercial basis rather than by legal factors. Dual residence does not arise in this case, and, in my judgment, as it has no bearing on the way I consider this appeal should be determined, I will say no more about it.[31]Mr Vallat submits that, although the Commonwealth cases on which Ardmore relied below are also based on legislation from different jurisdictions, the approach in them is still to look at a variety of factors; in any event they are not uniform in their approach. In particular, in Phillips, the New Zealand court followed Lever Bros, locating the transaction from which the interest took its origin. The Commonwealth cases considered below do not establish a consistent rule that the place of lending is a relevant factor in itself. Even if they did, the weight attached for the purposes of UK law must be minimal in light of National Bank Case, where it was not mentioned.[32]For the avoidance of doubt,(i) HMRC does not accept that the “nationality” of the document should be given weight as a separate factor and(ii) the irrelevance of situs extends to both the common law rule (applicable for income tax) and, a fortiori, the rule in TCGA 1992, section 275 (cited by Ardmore). As to the last point, I do not propose to say more about this argument than that section 275(1) is a specific statutory provision which in my judgment sheds no light on the present problem.[33]I now turn to my conclusions. The problem that has to be solved is this. It is now agreed on all sides that, following the National Bank case, the correct test is the multifactorial test but that necessarily throws up different indications as to different sources. In other words, the multifactorial test does not help the Court choose which the determinative factors are.[34]The answer to this particular problem is I think to be found in a number of passages in the decided cases, but I will choose the brief and elegant judgment of Lord Atkin, giving the advice of the Privy Council in Rhodesia Metals, which was cited to the FTT and to which I referred in paragraph 7 above. Lord Atkin first made the point, which I would respectfully adopt, that great caution should be applied in adopting decisions on source from different legislative regimes, that what is the source depends to some extent on the perspective from which the question is asked, that it was not possible to provide a universal definition of “source” and, most importantly, that the question had to be resolved by applying practical sense. I refer to this below as “the practical approach.”[35]The case concerned the source for the purposes of South African legislation of profits which an English company earned on selling in the UK mining claims which were immovable property in Zimbabwe. The Privy Council concluded that the source was in Zimbabwe (then Southern Rhodesia).[36]Lord Atkin referred to cases on the source principle and held: Their Lordships have no criticisms to make of any of those decisions, but they desire to point out that decisions on the words of one statute are seldom of value in deciding on different words in another statute, and that different business operations may give rise to different taxing results. If the charging words of the English statute are looked at, "(i.) annual profits or gains arising to any person, (ii.) residing in the United Kingdom from any trade wherever carried on, and (iii.) whether resident in the United Kingdom from any trade exercised within the United Kingdom"; they are obviously different from the Southern Rhodesian charging words, "total amount [other than capital] received by .... any person .... from any source within the Territory." It is desirable, also, to point out that, at any rate for different taxing systems, income can quite plainly be derived from more than one source even where the source is business. For instance, in the case of the business of a railway company whose railway is situate abroad, as in San Paulo (Brazilian) Ry. Co. v. Carter, while the English company may be assessed in England on the whole of its profits because it carries on part of its business there, yet it could not be doubted that so much of the profits of the business as were in fact earned from running the railway in Brazil were derived from exercising a business in Brazil; and still less could it be doubted that the sums received by the company in Brazil were received from a source in Brazil. In the present case their Lordships do not find it necessary to formulate a definition which will afford a universal test of when an amount is "received from a source within the territory." A doubt may be expressed whether the words borrowed by Stratford C.J. from Innes C.J. in the Overseas Trust case "productive employment of capital" really help to define the situation. Is capital productively employed in the place where it purchases stock which is profitably sold elsewhere? or in the place where the stock which now represents the capital is sold? or, for purposes of the test, must both purchases and sales occur in the same place? or is it sufficient that the place of the direction of the employment of the capital in purchasing or selling should denote where the capital is productively employed? Perhaps, in other words, it may be said, does it mean more than carrying on business in a place? Their Lordships incline to the view quoted with approval from Mr. Ingram's work on South African Income Tax Law by de Villiers J. in his dissenting judgment:
"Source means not a legal concept, but something which a practical man would regard as a real source of income"; "the ascertaining of the actual source is a practical hard matter of fact."
At any rate, in the present case, whatever may be the right view of the source of receipts derived from trading in commodities, their Lordships find themselves dealing with a case where the sole business operation of an English company is the purchase of immovable property in Southern Rhodesia and its development in that territory for purposes of transfer in that territory at a profitable price. The company never adventured any part of its capital except on that or those immovables. As a hard matter of fact the only proper conclusion appears to be that the company received the sum in question from a source within the territory, namely, the mining claims which they had acquired and developed there for the very purpose of obtaining the particular receipt. Their Lordships will therefore humbly advise His Majesty that this appeal should be dismissed. The appellant must pay the costs of the appeal. (paragraph breaks added)[37]The Privy Council’s reference to matters of fact means that the correct approach to applying the multifactorial test is to ask whether a practical person would regard the source as in this jurisdiction or elsewhere. I do not see any material difference between this approach and that of Lord Hailsham in the National Bank case. Lord Hailsham applied a matter of fact approach as opposed to an approach based on legal concepts and rules.[38]Lord Atkin’s approach has been impressed on us by HMRC, who have emphasised the need to have regard to “the underlying commercial reality.” Those words are appropriate to this case which concerns a commercial transaction but in other types of transaction one might ask which was the source from a practical, or realistic, point of view.[39]There has been much reference to a multifactorial test. To that I would add that the correct approach is the practical approach and that it is not merely multifactorial but also acutely fact-sensitive. The court or tribunal must examine all the available facts both singly and cumulatively.[40]That leads to the next point. The evaluative nature of the exercise of applying the source principle means also that the appellate tribunal should be slow to interfere, especially as in this case the Tribunals are specialist tribunals. Consequently, in my judgment, Ardmore has to satisfy this Court that the Tribunals were wrong in the sense that they left a material factor out of account or took a matter into account that should have been left out, or misdirected themselves in law or fact or reached a perverse conclusion.[41]I do not consider that Ardmore discharges this burden. It is sufficient for me to confine my examination to the evaluation of the Upper Tribunal though no doubt this was made with the benefit of the FTT’s analysis. The Upper Tribunal did hold that the lender’s place of residence and the place of credit were not relevant (and indeed HMRC urged on us that some factors were “irrelevant”) whereas on a multifactorial test a factor is still relevant even if it carries little weight.[42]In this case, however, the conclusion of the Upper Tribunal that the residence of the creditor should carry little weight cannot be criticised. The immediate search is for the source of the interest rather than a search indirectly for the source of the loan. The funds paid over as interest derived from funds generated in the UK. The activity of lending became passive once the loan was made, whereas the business of Ardmore was actively conducted to produce those funds. There was no default and the Gibraltarian exclusive jurisdiction and governing law clauses would only matter if there was default. The importance of those clauses is also undermined by the fact that the enforcement of any judgment following default on assets of Ardmore would be in the UK (and it is not necessary to go further than to note that all the available assets to meet the liabilities to the lender were in the UK). Furthermore, relative to the links with the UK, the links with Gibraltar were of an insubstantial kind: there was no evidence that they were backed up by any kind of other activity within Gibraltar, nor was it explained why it was necessary for the trusts to form companies in the BVI or what commercial purpose those companies served. The insubstantial nature of the transaction’s connection with Gibraltar distinguishes this case from the Philips and Rhodesia Metals cases where there was a significant connection with the source (active business and mining claims respectively). I see no basis, therefore, for holding that the Tribunals left out of account any material factor or took any immaterial factor into account.[43]Moreover, the Tribunals looked to the substantive matters rather than theoretical factors, such as causative link and governing law, and so applied a practical approach. It was as unnecessary to spell out that they were applying that approach any more than it had been spelt out by Lord Hailsham, who did not refer to it, but in fact followed it (see above, paragraphs 11 and 12). Furthermore, the decision in cases such as Philips cannot mandate the outcome in another case on different facts.[44]In these circumstances, in my judgment, there is no basis to which this court could properly interfere. I would dismiss this appeal.

LORD JUSTICE SALES:

[45]I agree.

LORD JUSTICE LEGGATT:

[46]I also agree.

Cited in 3 later judgments