This is an appeal from an order made on 16 June 2004 by Mr Justice Park in a test case brought under a group litigation order made following, and arising out of, the decision of the Court of Justice of the European Communities in the joined cases Metallgesellschaft Ltd v others v Inland Revenue Commissioners and another and Hoechst AG and another v Inland Revenue Commissioners and another (Cases C-397/98 and 410/98). The decision of the Court of Justice was delivered on 8 March 2001 and is reported at [2001] Ch 620.

The issue which was before the Court of Justice in the Metallgesellschaft/Hoechst cases was whether it was contrary to Community law - specifically, the provisions then contained in article 52 of the EC Treaty (now renumbered as article 43) - for the domestic tax law in the United Kingdom to differentiate, in the treatment of advance corporation tax on dividends paid by a subsidiary company to its parent, between cases where both subsidiary and parent were resident in the United Kingdom for tax purposes and cases where the subsidiary was resident in the United Kingdom but the parent was resident in another member state. The effect of the differential treatment under domestic law was that the subsidiary of a parent company which was resident in another member state suffered a timing disadvantage; in that it was required to pay an amount in respect of corporation tax earlier than it would have done had the parent, also, been resident in the United Kingdom.

The Court of Justice held that the differential treatment in relation to advance corporation tax was contrary to Community law. The Court went on to hold that a United Kingdom subsidiary of a parent company resident in another member state which had been required to pay, and had paid, an amount in respect of corporation tax earlier than would have been the case had its parent been resident in the United Kingdom was entitled to compensation for the timing disadvantage which it had suffered. The amount of that compensation was to be determined by the national court.

The group litigation order, which was made on 26 November 2001, identified a number of common or related issues of law arising out of the decision of the Court of Justice. Amongst those issues (at paragraph 9(ii) of the order, under the heading “EU Issues - Quantum”) were the following:

Those issues were raised – the first by an amendment made on 24 October 2003 - in the action which had been commenced by Metallgesellschaft Limited against the Commissioners of Inland Revenue as long ago as 1995. It was that action – and those issues – which were before Mr Justice Park in April 2004 as a test case in the group litigation. Since the issue of the writ the claimant had changed its name to the present name, Sempra Metals Limited, and the Attorney General had been joined as an additional defendant. The underlying question – as paragraph 8 of the skeleton argument prepared on behalf of the claimant made clear – was whether, in relation to each of the periods identified by the answers to issues (A) and (B), interest should be compounded. It was claimant’s case that:

In the judgment which he handed down on 16 June 2004, Mr Justice Park described the relevant period for the purposes of Issue (A) as “the premature tax payment period”. That period – in a case where an amount equal to the advance corporation tax actually paid by the subsidiary company was later set-off against a liability to corporation tax on taxable profits – was the period between the date of premature payment and the date of set-off. The judge decided that, in respect of that period, compensation (or restitution) should be calculated on the basis of compound interest. His decision on that point is reflected in paragraph 1 of his order dated 16 June 2004. It is from that paragraph (and the consequential order for costs in paragraph 5 of the order) that the Revenue appeals.

The judge described the relevant period for the purposes of Issue (B) – that is to say, the period from the date when an amount equal to the advance corporation tax paid was set-off against corporation tax on taxable profits – as “the post-utilisation period”. He held that, in respect of that period, the amount of compensation (or restitution) – whatever it might be – should bear simple interest. Effect is given to that decision by paragraph 2 of the order of 16 June 2004. There is no challenge to that part of the judge’s order.

Sempra Metals Limited (formerly Metallgesellschaft Limited, and referred to hereafter as “Sempra”) is a company resident in the United Kingdom. It trades, both as a principal and as a broker for clients, in metals listed on the London Metal Exchange. It was, at all relevant times, a subsidiary of Metallgesellschaft AG, a company having its seat in the Federal Republic of Germany. Over the years between 1974 and 1995 Sempra paid dividends to its German parent; and paid the associated advance corporation tax in respect of those dividends. It is accepted by the Revenue that, if the United Kingdom legislation had permitted group income elections to be made between United Kingdom subsidiaries and parent companies established in other member states, Sempra and its parent would have made such an election; with the consequence that Sempra would not have been required to pay advance corporation tax.

For the purposes of the test case, the parties identified and agreed four sample dividends, with associated payments of advance corporation tax. The earliest tax payment was made on 12 October 1981 and the latest on 18 July 1994. Each of the payments of advance corporation tax was subsequently set-off against corporation tax in respect of chargeable profits. The periods between payment and set-off (the “premature tax payment periods”) varied between one year and ten years. The judge found that, at almost all times, Sempra was in a net borrowing position, in the sense that its cash borrowings exceeded its cash assets.

The statutory regime under which advance corporation tax was payable on qualifying distributions made by a company resident in the United Kingdom is fully described at paragraphs [3] to [21] in the judgment of Lord Justice Peter Gibson in the first in the series of test cases to be tried under the group litigation order of 26 November 2001, Pirelli Cable Holding NV and others v Inland Revenue Commissioners [2003] EWCA Civ 1849, [2004] STC 130. It is unnecessary to rehearse every detail of that regime in this judgment.

It is enough, I think, to say this. Advance corporation tax was introduced in the Finance Act 1972 and remained payable until abolished, with effect from 6 April 1999, by the Finance Act 1998. From 1988 the relevant provisions were found in the Income and Corporation Taxes Act 1988 (“ICTA 1988”). The charging provision was contained in section 14(1) of that Act. The general rule was that advance corporation tax, when payable, was to be paid within 14 days after the end of the quarter in which the distribution was made – section 238(5) ICTA 1988, read with paragraphs 1(2) and 3(1) of schedule 13 to that Act. It could not be deducted from the amount of the distribution; but it could be set-off against the paying company’s liability to corporation tax in respect of chargeable profits (“mainstream corporation tax” or “MCT”) for the year of assessment within which the distribution was made; or (if the profits for that year were insufficient) against chargeable profits for future years of assessment – section 239 ICTA 1988.

In circumstances where the paying company had sufficient chargeable profits (in that or in future years) it would, ultimately, pay no more corporation tax than the amount which it would otherwise have paid as MCT. But the effect on the paying company, as Lord Justice Peter Gibson explained at paragraph [8] of his judgment in Pirelli Cable Holding NV (supra), was that the date for payment of the corresponding MCT which would otherwise be due was advanced by a period of between eight and a half months (if the distribution was made on the last day of an accounting period) and twenty nine and a half months (if the distribution was made on the first day of an accounting period). So there was a ‘cost’ to the paying company, in the sense that it had lost the use of money (in the amount of the advance corporation tax which it had paid) for a period which was not insubstantial.

The charging provision, section 14(1) ICTA 1988, was expressed to be subject to section 247 of that Act. Section 247(1) provided that where a parent company received a dividend from a 51 per cent subsidiary – and both parent and subsidiary were resident in the United Kingdom – the receiving company and the paying company could join in making a group income election. The effect of a joint income election was that dividends received by the parent from the subsidiary were excluded from section 14(1) of the Act. They were treated as “group income” of the parent for the purposes of the Corporation Tax Acts. The paying company was spared the ‘cost’ of losing the use of money between the date when it would have paid advance corporation tax and the date of set-off.

It was the requirement, in section 247(1) ICTA 1988, that the receiving company, as well as the paying company, should be resident in the United Kingdom that led to differential treatment, in relation to advance corporation tax, between a case where both parent and subsidiary were resident in the United Kingdom and a case where the subsidiary was resident in the United Kingdom but the parent was not so resident. In the former case, the subsidiary was spared the ‘cost’ of losing the use of money between the date when it would have paid advance corporation tax and the date of set-off; in the latter case the subsidiary had to bear that cost. The effect of the differential treatment was that the subsidiary of a parent company which was resident in another Member State was required to pay an amount in respect of corporation tax earlier than it would have done had the parent, also, been resident in the United Kingdom. The subsidiary company suffered a timing disadvantage in that respect.

The question whether it was consistent with Community law – and, in particular, consistent with what was then article 52 of the EC Treaty - for the legislation of a member state to permit a group income election (allowing distributions to be paid by a subsidiary to its parent without accounting for advance corporation tax) only where both the subsidiary and the parent were resident in that state was referred to the Court of Justice in the course of the present proceedings. The Court answered that question in the negative. It held that:

That answer led to a second question:

The Court of Justice answered that question in the affirmative:

It can be seen that the question posed – and the answer – address the possibility that, at the date when a remedy is sought, it will be known that there has been no overpayment of corporation tax – because, by that date, the advance corporation tax paid prematurely has been set-off against the paying company’s liability to corporation tax in respect of chargeable profits. But, as the Court of Justice makes clear “The mere fact that the sole object of such an action is the payment of interest equivalent to the financial loss suffered as a result of the loss of the use of the sums paid prematurely does not constitute a ground for dismissing such an action”. The rule of domestic law, founded on the decision of the House of Lords in London, Chatham and Dover Railway Co v South Eastern Railway Co [1893] AC 429 and upheld in President of India v La Pintada Cia Navigacion SA [1985] AC 104, which precludes an action at common law for an award of interest by way of damages for late payment of a debt could not be invoked to deny an effective legal remedy in respect of the financial loss which had been sustained by reason of the premature payment of advance corporation tax.

The point is, perhaps, put most clearly in the opinion of the Advocate General (Fennelly), at paragraph 46 ([2001] Ch 620, 642):

It can be seen, also, that the second question referred to the Court of Justice was posed on the basis of alternative hypotheses. That that was how it appeared to the Court is made clear at paragraph 77 of the judgment (ibid, 662C): “In the first alternative, where the claim by the subsidiary and/or parent company is made in an action for restitution of taxes levied in breach of Community law and, in the second where the claim is made in an action for compensation for damage resulting from the breach of Community law”. The Court approached the question on the basis that it was not its role to assign a legal classification to the action to be brought by the claimants before the national court: “. . . it is for the claimants to specify the nature and basis of their actions (whether they are actions for restitution or actions for compensation for damage) subject to the supervision of the national court” – (ibid, 663A-B, at paragraph 81).

The Court of Justice did not find it necessary to decide whether the national court would classify the claim as restitutionary or as compensatory – although the Advocate General had found it “more correct and more logical to treat the claimants’ claim as restitutionary rather than as a compensatory claim for damages” – (ibid, 645H, paragraph 52). In the Court’s judgment the answer to the second question was the same, whichever classification was adopted. That appears from paragraphs 82 to 89 (ibid, 663B-664E) – where the question is addressed on the basis that the claim is restitutionary – and from paragraphs 90 to 95 (ibid, 664F-665D) – where it is addressed on the basis that it is a compensatory claim. It is, I think, important to understand why the Court of Justice reached that conclusion.

The Court defined the question (on the first hypothesis) at paragraph 82:

At paragraphs 85 and 86 of its judgment the Court acknowledged (i) that, in the absence of Community rules on the restitution of national charges that had been improperly levied, it was for the domestic legal system to lay down detailed procedural rules governing actions for safeguarding rights which individuals derive from Community law - subject to the principle of equivalence and the principle of effectiveness – and (ii) that it was for national law to settle all ancillary questions relating to the reimbursement of charges improperly levied “such as the payment of interest, the rate of interest and the date from which it must be calculated”. But the Court went on to say this, at paragraphs 87 to 89:

Turning to consider the question on the alternative hypothesis, the Court of Justice said this, at paragraph 90 of its judgment (ibid, 664F):

I have set out the reasoning of the Court of Justice at some length in order to give emphasis to the two factors which, as it seems to me, determine the outcome of this appeal. First, the national court is required to give a remedy, whether by way of restitution or as compensation, in respect of the breach of Community law. It is not open to the national court to deny restitution or compensation on the ground that no remedy would lie under domestic law. If necessary, Community law demands an autonomous remedy in respect of the breach of Community law which has occurred. Second, the remedy to be given by the national court must be a “full” remedy; in the sense that it must be such as will restore the equality of treatment guaranteed by article 52 (now article 43) of the EC Treaty. Nothing less will do. A full remedy for the loss of the use of money over a specified period may be measured by reference to the interest “accrued” on the amount of the tax paid prematurely. But it is important to keep in mind that there is no true analogy with the award of interest on a domestic judgment. The task of the national court is to ascertain the amount which the member state must pay to the claimant in order to restore the claimant to the position that it would have been in if it had not been required to pay an amount of corporation tax prematurely.

At paragraph 16 of his judgment, the judge set out, under six sub-paragraphs, propositions which he derived from paragraphs 77 to 96 of the judgment of the Court of Justice. I need not rehearse those propositions verbatim. For the most part, they are already included in the analysis of the judgment of the Court of Justice which I have, myself, set out in the preceding section of this judgment. It is enough to say that I agree with the judge’s summary. At paragraph 17 of his judgment the judge observed that, although the Court of Justice referred to “interest” and the period in relation to which it should be computed (the premature tax payment period), the Court did not, in terms, express a view whether interest over that period should be computed on a simple basis or on a compound basis. He saw it “as my principal task” to decide which of those two bases – simple or compound interest – “appears more precisely to accord with the principles which underlie the court’s decision”.

Before addressing that question, the judge asked himself whether the amount of the restitution or compensation should be calculated on a “conventional basis” – that is to say, by reference to interest and at a rate of interest which is the same for all claimants – or on an “actual basis”, by examining the particular circumstances of each claimant company. He held that it was appropriate to adopt a conventional basis in all cases. He thought it implicit in the decision of the Court of Justice that a claim to be recompensed for loss of the use of money was to be measured by reference to interest – rather than by an enquiry into the return on the money which the claimant would have obtained if it had not been required to make premature payment. And, at paragraph 22 of his judgment, the judge said this:

The judge held that the computation should be on the basis that interest was compounded. His reasoning is encapsulated in paragraphs 25 to 27 of his judgment:

The appellants submit that the judge was wrong to approach his task on the basis that he had to “look for the measure of restitution or compensation which Community law requires, untrammelled by restrictions which English law may impose in situations which might be said to be comparable in some respects”. It is said that that is not what Community law requires. Properly understood, the decision of the Court of Justice was that “where the loss to be made good is the time value of money, . . . full compensation is given by an award of interest, calculated in accordance with the normal national approach”. In awarding compound interest, the judge departed from the normal approach under English domestic law.

In support of the proposition that the Court of Justice had in mind, and was content, that an award based on simple interest would meet the need to provide full compensation for the loss of the use of money in a case where, in breach of Community law, the claimant had been required to make a premature payment of tax the appellants relied on the decision of the Court in Marshall v Southampton and South West Hampshire Area Health Authority (Teaching) (No 2) (Case C-271/91) [1994] QB 126. The appellants pointed out that the Court of Justice had referred to its decision in Marshall at paragraph 94 of the judgment in Metallgesellschaft/Hoechst ([2001] Ch 620, 665B-D) :

It is pertinent to have in mind the underlying facts in the Marshall case. Mrs Marshall had been dismissed by her employer, the health authority, at the age of 62 years, on the sole ground that she had reached the retirement age for female employees stipulated by the authority. She claimed that, since the stipulated retirement age for male employees was 65 years, she had suffered unlawful discrimination on the grounds of her sex. That claim was upheld by the Court of Justice in Marshall v Southampton and South West Hampshire Area Health Authority (Teaching) (Case 152/84) [1986] QB 401. Her claim was remitted to the industrial tribunal (as it then was) to assess compensation. The tribunal awarded a sum which included £7,710 in respect of interest. It did so in purported reliance on section 35A of the Supreme Court Act 1981. That award was challenged by the health authority on the ground that the tribunal had no power, under domestic law, to award interest. That challenge was upheld by the appeal tribunal and by this Court. On a further appeal to the House of Lords, the question was referred to the Court of Justice for a preliminary ruling. At paragraph 31 of its judgment ([1994] QB 126, 165F) the Court gave the ruling to which it was later to refer in paragraph 94 of the judgment in Metallgesellschaft/Hoechst (supra).

It is submitted, on the present appeal, that the interest awarded to Mrs Marshall by the industrial tribunal must have been simple and not compound interest. The award had been made under section 35A of the Supreme Court Act 1981; and simple interest is all that that section 35A(1) allows:

That, as it seems to me, is to treat the decision in the Marshall case as determinative of an issue which was never before the Court of Justice and to which the Court never turned its mind. The Court was concerned only with the question whether, in principle, full compensation for loss suffered as a result of the discriminatory dismissal could leave out of account a factor – the effluxion of time – which might, in fact, reduce its value. It decided that the award of interest, in accordance with applicable national rules, was an essential component of compensation for the purposes of restoring real equality of treatment. But there is nothing to suggest that the Court was asked to consider – or did consider – whether the amount of interest actually awarded in the Marshall case was sufficient for those purposes. I am not persuaded that the Marshall case provides any support for the proposition that the Court of Justice was content to assume, in the present case, that an award based on simple interest would meet the need to provide full compensation for the loss of the use of money which the claimant had been required to pay prematurely.

The appellants rely, also, on the decision of the Court of First Instance of the European Communities in Corus UK Ltd v Commission of the European Communities (Case T-171/99) [2002] 1 WLR 970. The underlying facts may be stated shortly. In February 1994 the Commission had found that Corus and other steel undertakings had participated in a series of anti-competitive agreements and practices on the Community steel beam market, in breach of article 65(1) of the ECSC Treaty. The Commission imposed a fine, which Corus paid in June 1994. Subsequently, in March 1999, the Court of First Instance annulled the February 1994 decision in part, and reduced the amount of the fine. On 23 April 1999 the Commission repaid €12m. Corus sought interest on that sum in respect of the period from June 1994 to April 1999. The response of the Commission was that it had satisfied its obligations under article 34 CS by repaying the principal sum. The article was in these terms:

The Court rejected the claim in so far as it was based upon the first paragraph of article 34: faults of such a nature as to render the Community liable, or direct and special harm. But it went on to say this, at paragraph 49 of its decision (ibid, 979E-F):

The Court held, at paragraph 53 of its decision, that the obligation to repay extended not only to the principal amount of the fine overpaid, but also to ‘default interest’ on that amount. At paragraph 54 the court referred to the principle set out at paragraph 31 in Marshall (No 2) and at paragraphs 94 and 95 in Metallgesellschaft/Hoechst: “complete reimbursement of a fine unduly paid cannot leave out of account factors, such as the effluxion of time, which may in fact reduce its value . . . Proper compliance with such a judgment therefore requires, in order fully to restore the applicant to the position in which it legally should have been, that it be taken into account that such restoration only occurred after an appreciable lapse of time, during which the applicant did not have the use of the sums it had unduly paid . . .”.

At paragraphs 60 to 63 of its decision in the Corus case, the Court of First Instance addressed, directly, the question what amount should be allowed in respect of the loss of the use of the money that had been overpaid. Before setting out those paragraphs it is convenient to note two matters. First, Corus assessed the cost of its loss of the use of the €12m overpaid by reference to the interest lost to it resulting from the diminution of its cash balances in respect of the sterling equivalent of that sum. It did so on the basis that, during the relevant period, it was “in a cash surplus situation and was investing funds on a three month rolling basis with interest compounded”. The lost earnings from investment were put at £3,533,474 in respect of the period June 1994 to April 1999 (paragraph 17). Second, there was evidence from the Commission that fines were added to levies and invested, and so earned interest for the Community (paragraph 31). The yield over the period, at an average rate of 4.613%, “taking account of the quarterly capitalisation of interest”, was calculated at €3,016,608 (paragraph 32). With those matters in mind, the Court of First Instance said this:

The appellants rely, of course, on the observation, in paragraph 61 of the judgment of the Court of First Instance, that “it would normally be appropriate to award the applicant simple interest on the sum of €12m, at a fixed rate to be determined by the court, for the period from 2 June 1994 to 23 April 1999”. For my part, I do not find that observation of much assistance; when set in the context of what the court did in fact. Although the court explained that “according to a principle generally accepted in the domestic law of member states . . . the claimant is normally entitled to the lower of the two amounts corresponding to the enrichment and the loss”, it made no calculation of what the loss would have been on the basis of simple interest at a fixed rate. And it is not self-evident that loss calculated on the basis of simple interest over the period of nearly five years (Footnote: ¹) would not have been less than the amount actually awarded – which was, itself, based on compounding interest. Mr Justice Park took the view that no general principle could be extracted from the Corus case. I think he was correct to take that view.

The more powerful argument advanced on behalf of the appellants, as it seems to me, is that based on the anomaly which, it is said, would arise between what the appellants describe as ‘category 1’ cases and ‘category 2’ cases if the judge’s view were to prevail. In that context, a category 1 case is one (such as the present) in which the United Kingdom subsidiary has sufficient chargeable profits in the year of assessment (or in subsequent years) to be able to set-off advance corporation tax against corporation tax on chargeable profits. That is to say, the subsidiary has been able to utilise the advance corporation tax paid. A category 2 case is one in which the subsidiary has not been able to set-off advance corporation tax (or all advance corporation tax) paid against corporation tax on chargeable profits; and so has become entitled (in the light of the decision in Metallgesellschaft/Hoechst) to a repayment of advance corporation tax. It is said that, under domestic law, the claimant in a category 2 case would be entitled to recover the overpaid advance corporation tax as a principal sum, together with simple interest on that sum under section 35A of the Supreme Court Act 1981, from the date on which the advance corporation tax was paid until the date of judgment. So, it is submitted, “the remedy in category 1 claims should be that the claimant is entitled, as a basic sum, to simple interest on his ACT for the premature tax period. Only in that way can the taxpayers whose ACT payments have been set off be treated in the same way as those still left with surplus ACT”. The anomaly, if it exists, is presented starkly in a case where the claimant has been able to set off against corporation tax on chargeable profits some, but not all, of the advance corporation tax paid prematurely. In such a case the claimant has claims (i) for interest on the amount of advance corporation tax paid prematurely but subsequently set-off (for the period from the date of payment to the date of set-off), (ii) for repayment of the amount of the advance corporation tax which it has not been able to set-off, and (iii) for interest on the amount of the advance corporation tax paid and not subsequently set-off – that is to say, for interest on the amount of the repayment – for the period from the date of payment to the date of judgment. It would be strange if interest under claim (i) were calculated on a different basis from interest under claim (iii) – the one compound and the other simple - in circumstances in which the period for interest under claim (i) would comprise part (at least) of the period for interest under claim (iii).

The judge addressed that argument at paragraphs 33 to 39 of his judgment. Put shortly, he took the view that, if there were a true anomaly between the treatment of category 1 and category 2 cases, it was the treatment of category 2 cases under domestic law which was impossible to reconcile with what Community law required; those cases should be seen as exceptional cases; and the anomalous treatment did not provide a sufficient reason to refuse to give effect to the requirements of Community law in the, more normal, category 1 cases. But he went on to say this, at paragraph 37):

It is important not to lose sight of the fact that, in a case (such as the present) where there is no principal sum outstanding at the time when the action is commenced, a claim for interest by way of restitution or compensation for loss of the use of money during a period over which the principal sum was outstanding would not be allowed under the domestic law of the United Kingdom – President of India v La Pintada Compania Navigacion SA [1985] 1 AC 429. So, as the judge held (at paragraphs 24 and 25 of his judgment), the domestic rules in relation to interest are unlikely to be of much (if any) assistance. The task for the English court is to give the remedy which Community law requires in circumstances where domestic law would not provide a remedy.

Nevertheless it is, I think, pertinent to have in mind that – as the Law Commission observed at paragraph 2.3 of its recent report, Pre-Judgment Interest on Debts and Damages” (Law Com No 287):

It is, I think, generally accepted that, in commercial cases at least, an award based on compound, rather than simple, interest is more likely to reflect economic reality and to provide full, rather than partial, recompense for the loss of the use of money. That was the view of the Law Commission, which (in the report to which I have referred) recommended that the courts be given power to award compound interest. It was the view expressed in trenchant terms by Mr Justice Hobhouse in Westdeutsche Landesbank Girozentrale v Islington London

The reason why compound interest is required to achieve full restitution or recompense is to be found in the basis upon which interest rates are set (or agreed) in commercial transactions. It is necessary to have in mind that parties to a commercial loan expect interest to be paid (or accrued) periodically. The rate of interest is set on the basis that the interest will be paid (or accrued) at the end of each period. So a bank may lend at a daily rate, a monthly rate or a quarterly rate, or perhaps at a half-yearly or yearly rate. In each case the rate may be expressed as an annual rate - say, 5% per annum. But the return on a loan at 5% per annum will vary according to the periodic basis (daily, monthly, quarterly, etc.) on which that rate has been set.

The point may be illustrated by an example. A bank which lends £10,000 on 1 January at a rate of 5% per annum payable quarterly will expect to receive £125 on each of 1 April, 1 July and 1 October. And, if it does receive those interest payments on the due dates and re-lends them on the same terms, it will have received £509.45 by 1 January in the following year (Footnote: ²). In order to obtain the same return over one year on a loan at simple interest, the rate of interest would need to be 5.0945%. Or, to put the point another way, if the rate of interest chosen is the rate appropriate to a commercial loan (say, 5% per annum fixed on a quarterly basis), but only simple interest is allowed, the return at the end of the year will be £9.45 less than it would have been done if the loan had been made on commercial terms.

There are two other factors which should be kept in mind. First, the ostensible object of an award of interest under section 35A of the Supreme Court Act 1981 is to compensate the claimant for the loss of the use of money. An award under that section is not made in order to prevent the unjust enrichment of the defendant – although it may have that effect. That is in contrast to the basis upon which, under domestic law, compound interest may be awarded in equity. The position is explained by Lord Browne-Wilkinson in the Westdeutsche Bank case (ibid, 700H-702E). After referring to the statements of principle by Lord Hatherley, Lord Chancellor, in Burdick v Garrick (1870) LR 5 Ch App 233, 241, and Lord Justice Buckley in Wallersteiner v Moir (No 2) [1975] QB 373, 397, and to observations of Lord Brandon of Oakbrook in President of India v La Pintada Compania Navigacion SA [1985] AC 104, 116, Lord Browne-Wilkinson said this:

Second, the convention – or, at least, the practice in the Commercial Court - when awarding interest under section 35A of the Supreme Court Act 1981, is to treat the claimant as a net borrower; and to award interest at a rate which reflects the cost of borrowing rather than the return on lending. The point was addressed by Mr Justice Forbes in Tate & Lyle Food & Distribution Ltd v Greater London Council [1982] 1 WLR 149, 154. His conclusion – that a borrower’s rate should be awarded – was adopted by Mr Justice Webster in Shearson Lehman Hutton Inc and another v McLaine Watson & Co Ltd and others (No 2) [1990] 1 Lloyd’s Rep 441, 451-453. That approach was followed in this Court in Saleem Jaura v Saeeda Ahmed [2002] EWCA Civ 210 (unreported, 21 February 2002). In that latter case this Court was ready to accept that the law should recognise that the rate of interest at which a small business man could borrow was likely to be higher than that available to a first class institutional borrower.

I mention those points in order to put in context the observations of the Court of First Instance in the Corus case – to which I have referred earlier in this judgment. The award actually made in that case was comparable to that which could have been made in equity, under English domestic law, in a case where the defendant was a defaulting trustee – but not otherwise (see the decision in the Westdeutsche Bank case). The award which was sought by the claimant in Corus was based on a lender’s rates – because the claimant accepted that it was in cash surplus. It is not at all clear that the Court of First Instance appreciated that – under English domestic law, at least – it is unlikely that either of those two alternatives would have been adopted. But that, perhaps, serves only to illustrate that, in order to provide the remedy which Community law requires, the national court may find little or no assistance in its own domestic rules as to interest.

In my view the judge was correct to hold that the task of the national court was to give effect to the decision of the Court of Justice by providing the remedy in respect of the loss suffered by the taxpayer by reason of the premature payment of advance corporation tax which Community law requires. That remedy is full compensation for the loss of the use of money. The measure of compensation is interest accrued on the money over the premature payment period.

The judge was invited to approach the task on the basis that the rate of interest to be applied should be a conventional rate; that is to say, that he should not seek to tailor the rate to the individual circumstances of each taxpayer. And, if the matter is to be approached on the basis of “one rate suits all”, then (as it seems to me) there is really no alternative to adopting a borrower’s rate rather than a lender’s rate. To adopt a lender’s rate in all cases would lead, inevitably, to under-compensation in those cases where the taxpayer was in cash deficit. To adopt a borrower’s rate will (or may) lead to over-compensation where the taxpayer is in cash surplus; but that, as I understand the Revenue’s position, is accepted as price worth paying for the practical advantages of a single conventional rate.

Accepting, therefore, that compensation is to be measured by interest accrued at a single conventional rate – which is to be a borrower’s rate – the question is whether the computation should be made on the basis that interest is compounded at periodic intervals.

The judge thought that only compound interest would fully restore equal treatment. The premise which underlies that conclusion is that the rate chosen will be a rate which has been fixed by reference to the rates at which commercial loans are offered in the market – for example, London Inter-Bank Offered Rate (LIBOR) or the base rate of one or more of the London Clearing Banks. And, as I have said, the rates at which commercial loans are offered are set on the basis that interest will be paid periodically. So, if the rate chosen is set by reference to the rates at which commercial loans are offered on the market, a computation made on the basis of simple interest at that rate will be made on a false basis. It will ignore a critical feature which is inherent in rate which has been chosen. And, save in the rare case where the premature payment period is less than the period by reference to which the rate is set, the computation made on the basis of simple interest will produce an undervalue.

For those reasons I think that the judge was right to reach the conclusion that he did. Community law requires full compensation for the loss of the use of money; and full compensation for the loss of the use of money requires that interest is compounded – at least where the rate of interest chosen is a rate set by reference to the periodic payment or compounding of interest. The English domestic rules as to interest fail to provide the remedy which Community law requires. Those rules must yield to the overriding requirement that the domestic court gives full compensation. And, to my mind, the judge was correct to take the provisional view – expressed at paragraph 37 of his judgment - that that applies as much to unutilised ACT as it does to utilised ACT

For my part, I would vary paragraph 1 of the order of 16 June 2004 so as make it clear that the rate of interest adopted dictates both the principle that interest should be compounded and the periods (or rests) at which it should be compounded. Subject to the observations of counsel, I would declare that “Interest over the premature tax period (as defined in paragraph 6 of the judgment of Mr Justice Park) should be computed by compounding at the same periodic rests as those by reference to which the applicable rate has been fixed.” Save for that variation to paragraph 1 of the order, I would dismiss this appeal.

I agree with Lord Justice Chadwick’s judgment, including the variation he would make to the order, and have nothing to add.

I also agree.