HM Revenue and Customs v Marks and Spencer Plc

These applications for permission and appeals by both HMRC (the Revenue) and by Marks and Spencer PLC (M&S) concern claims for group relief in respect of losses of Marks & Spencer (Deutschland) GmbH (MSG), resident in Germany and Marks & Spencer (Belgium) NV (MSB), resident in Belgium. They are appeals from decisions of the Upper Tribunal on 21 June 2010 as to the availability of cross-border group relief and as to the method of quantifying such relief as was available.

A Member State's right to preclude group relief of losses sustained by non-resident subsidiaries and the circumstances in which such a prohibition infringes the right to freedom of establishment and of movement enshrined in Articles 43 and 49 TFEU (ex Arts 43 and 48 EC) are not new issues. Those issues have occupied the Tribunals of this country and the European Court of Justice (now the Court of Justice of the European Union) (together the ECJ)) since 2002. Despite consideration by Special Commissioners, by Park J, who referred preliminary questions to the ECJ, the decision of the ECJ (Case C-446/03 Marks & Spencer plc v Halsey, 13 December 2005 [2006] Ch 184), the judgment of Park J, analysing that decision [2006] EWHC 811 (Ch), and of the Court of Appeal [2007] EWCA Civ 117 (the First Appeal) (all of which judgments I will refer to as M&S v Halsey), and the decisions of the Tax Chamber (the FTT) on 2 April 2009 and 24 August 2009 and of the Upper Tribunal, those remain unresolved. Not even the judgment of this court, second time around, will provide quietus. There remain other issues to be decided and the likelihood of other appeals. New readers, if any remain, may learn the facts from the detailed findings of the Tax Chamber (FTT) and [6]-[12] of the decision of the Upper Tribunal. The decision of the ECJ has been the subject of detailed analysis by Park J and by this court in the First Appeal. I do not propose to repeat the facts, save where necessary to explain my view, or to set out, yet again, lengthy quotations. I shall assume that there are no new readers but that they have grown old in their pursuit of finality. I shall identify in respect of which issues permission to appeal is required, and who is seeking to appeal when I come to a particular issue.

MSG ceased trading in August 2001 and was dissolved following liquidation on 14 December 2007. MSB ceased trading on 22 December 2001 and was dissolved following liquidation on 27 December 2007. M&S’s claims to group relief in respect of MSG’s losses for 1996 and 1997, and for MSB’s for 1998 and 1999 were governed by the corporation tax pay and file rules in Schedule 17A to the Taxes Act 1988. The FTT dismissed its original claims on the grounds that the ECJ's paragraph 55 criteria (called the 'no possibilities test') were not satisfied and that second and subsequent claims were out of time. M&S assert that those subsequent claims were, by operation of the principle of effectiveness, in time. Claims were made for MSG losses from 1996 to 2002 and for MSB losses for 1999 to 2001 (depending on which of the six methods of calculation adopted, losses arise in different periods). The first group relief claims were made at a time when neither subsidiary was in liquidation, though in the period when they were made, between 2000 and 2003, both had ceased trading. M&S announced the closure of operations on 29 March 2001. MSG ceased trading in August 2001. MSB ceased trading on 22 December 2001. The second claims, on 20 March 2007, were made when both companies were in liquidation, the third, on 12 December 2007, just before both were dissolved, and the fourth, on 11 June 2008, on behalf of MSB following MSB's dissolution. The claims for years from 2000 onwards were governed by the self-assessment rules in Schedule 18 to the Finance Act 1998. They were within statutory time limits, whereas the second to fourth claims in respect of ‘pay and file’ years were, absent M&S's current contention as to effectiveness, outside statutory time limits.

The following issues arise :-

The Court decided that the UK was entitled to preclude relief for losses sustained in another Member State, save only in circumstances where those losses could not be used in that Member State. Analysis, yet again, of this decision is important. The decision of the ECJ established, so the Revenue contend, the lawfulness of the UK's restriction and left no room for consideration of any circumstances other than those at the time the losses were crystallised; it admitted of no relief other than where those circumstances, objectively judged, demonstrated that losses could not have been utilised in the past and could not be used in the future.  In particular, it left no room for an assessment of the individual circumstances of the surrendering company in the future, still less for its motives.

Even if that was not decided by the ECJ in M&S v Halsey, the Revenue submits that that was certainly decided by the ECJ in subsequent cases, thus permitting this court to adopt that approach and follow those cases, even if the Court of Appeal previously decided to the contrary.

The ECJ answered the first part of question 1.  It focussed on the group relief provisions within ICTA 1988, which before 2000 restricted losses available for group relief provided in ss 402 and 403 to those surrendered by companies resident in the United Kingdom and, after 2000, to profits and losses within the scope of United Kingdom tax law [17].  The Revenue contend that the key to the ECJ's decision lies in appreciating that it was concerned only with the lawfulness of the UK provisions; the Court was concerned only to determine whether Articles 43 and 48 precluded the restrictions within the United Kingdom group relief provisions.  The Court decided that those Articles did not preclude the United Kingdom restrictions save in circumstances where the non-resident subsidiary had exhausted the possibilities of using its losses for previous and future accounting periods in the state where it resided.

The Revenue submit that once the losses of a non-resident subsidiary are crystallised in an accounting period, the only question is whether it is impossible to use those losses in that foreign state.  This question should not be answered by reference to any action taken by M&S’s subsidiaries in Belgium or Germany after the end of accounting periods in which the losses were suffered.  What was required was a determination, once a claim had been made, as to whether the losses claimed came within a category in respect of which a cross-border restriction was compatible with Art 43 or whether, because it was impossible to have used them in previous accounting periods or to use them in future periods, they fell within a category in respect of which the UK was required to permit cross-border relief.

Resolution of many of the issues in this appeal will depend upon whether it is open to this court, in the light of its previous decision, to accept that crucial submission, and if we are not bound by that decision, whether that submission is a correct reading of the judgment of the ECJ.

That the judgment of the ECJ has been subjected to repeated analysis in UK courts and tribunals, may induce Welschmerz, but is not surprising. The Court does not appear to have chosen between rival arguments.  Those arguments adopted extreme positions: the Commission regarded the absence of provision for trans-national relief as a major obstacle to the proper functioning of the internal market (see [3] and [fn 6] in the opinion of Adv. Gen. Poiares Maduro) and the United Kingdom argued that such relief interfered with a Member State’s competence and control over its tax system.  It relied upon the fiscal principle of territoriality and the necessity to ensure cohesion within the UK tax system ([57] A-G’s opinion). Thus there were no submissions in relation to the adoption of the criteria on which the ECJ alighted at paragraph 55, still less as to how they should be applied. The debate was not concerned with the central problem in the instant appeal, namely how the Tribunals and Courts of Member States should apply those criteria.  The issues which have exercised domestic tribunals and courts since then were not aired.  I venture to observe that the Advocate General, in proposing a solution which was the precursor to the court’s decision, acknowledged that it might appear complex [83].

The Advocate General proposed that foreign losses should be afforded equivalent treatment to that applicable to the losses of resident subsidiaries [76] and [77].  But in answer to the Netherlands' objection that to do so would cause general disruption to national systems by allowing companies to transfer losses to those companies resident in states with higher rates of taxation, described as ‘trafficking in losses’, he proposed a solution which is echoed in the paragraph 55 test, that ability to transfer losses would not give rise to trafficking in losses in circumstances where :

“the losses of foreign subsidiaries cannot receive advantageous tax treatment in the State in which those subsidiaries are resident.  Where the State in which the foreign subsidiaries are established enables those subsidiaries to impute their losses to another person or to carry them forward to other financial years, the United Kingdom is entitled to oppose a claim for the trans-national transfer of those losses.[79] (his emphasis)”.

The reference to a state enabling the use of losses seems to me to be a reference to the state tax regime and not to the particular facts of the subsidiary’s case.

The Advocate General sought to meet the objection that it would be ‘excessively difficult’ for the United Kingdom to ascertain whether there was a possibility of group relief in another State by referring to the availability of mutual assistance and the power of a Member State to impose duties to give information as to the tax situation in the foreign country, provided that those requirements did not exceed the objective for which the information was sought [81].

The Advocate General’s proposals clearly resonated with the Court.  It too rejected the submission of the United Kingdom and of other Member States that the fact that they did not tax the profits of non-resident subsidiaries itself justified restricting group relief to losses surrendered by resident subsidiaries (Judgment [40]). The Court rejected the argument that a mere reduction in tax revenue, resulting from freedom to surrender losses cross-border, justified a measure which may inhibit freedom of establishment [44].   But the Court was concerned with the implications of an unconditional freedom to surrender losses across national borders [42].

It did recognise that to allow companies the option to choose in which Member State losses should be relieved would damage the balanced allocation of power to impose taxes.  That was the very problem identified by the Netherlands as liable to cause general disruption to national tax systems, described as ‘trafficking in losses’.  The criteria laid down by the ECJ in paragraph 55 were designed to prevent the damage it was perceived might be caused should companies be permitted to choose the most favourable tax regime in which to relieve their losses. The paragraph 55 criteria seek to identify circumstances in which there exists no possibility of choice.

It is important, for the purposes of this appeal to keep in mind that the conditions which must be fulfilled before it can be said that the restriction is contrary to Articles 43 and 48 are conditions designed to meet not merely the danger of the double use of losses [47] and [48] but also the risk of disruption whenever companies are permitted to choose the State in which they relieve their losses.  The conditions in paragraph 55 can only be fulfilled when it can be demonstrated that companies have no choice as to the country in which losses are relieved.

The issue both Park J and the Court of Appeal had to resolve was whether the facts by reference to which the paragraph 55 conditions had to be satisfied were those which existed or could be foreseen at the end of the accounting period in which the losses crystallised or at the date of the claim (CA [32]).

Park J took the view that what he called ‘the relevant time’ was the time when M&S made its claim for group relief.  His reasoning is, of course, instructive and illuminating.  The end of the accounting period was “too soon and likely to rule out virtually every case”  because at that time the resident subsidiary “was still likely to be carrying on its trade” and the law of the state of residence was likely to provide for the possibility of relief for the losses.[44].

It is of importance to recall Park J's reason for rejecting the date as the date of the hearing of the appeal.   If the relevant time was when the issue, as to whether the paragraph 55 conditions were satisfied, was decided by the fact-finding tribunal, it would permit a group  to “spin out time before the matter came to appeal in the hope that by then the facts would change “.[45]

He concluded that the relevant time was when the claim was made:

“If a company claims group relief at a time when the paragraph 55 criteria are satisfied it should get the relief.  If it applies for it at a time when the criteria are not satisfied it should not.”[46]

Chadwick LJ, with whom the other members of the court agreed, acknowledged that if the paragraph 55 conditions were satisfied, “full effect has been given to the power of the state of residence of the surrendering company and there is no need to ‘protect a balanced allocation of the power to impose taxes’” and “there is no potential for tax avoidance - the surrendering company has been required to use its losses in its own state of residence so far as it can.” [35]. But he took the view that that rationale for the paragraph 55 conditions did not justify the proposition that those criteria must be satisfied at the end of the period in which the surrendering company’s losses accrued. The question whether the restriction is lawful only arises when a claim is made and he concluded:

"The question whether the United Kingdom tax authorities are precluded by Community law from applying the restriction on group relief turns on whether the paragraph 55 conditions are satisfied. I can see no reason in principle why the latter question - whether the paragraph 55 conditions are satisfied -should not be answered by reference to the facts as they are when the former question arises” [36].

He found support in the fact that a resident company can choose to surrender its losses at or up to the time when a claim is made for group relief and saw no reason for differential treatment in the case of a non-resident's surrender [41]. He saw no potential for abuse in ruling that the decision whether the paragraph 55 conditions are satisfied may be based on facts as they are at the date of claim [42] (he twice referred to that date as “the end of the period within which the claimant company is permitted to make its claim”, but, must, I think, be referring to the date of a claim where such a claim has been made).

It seems to me that the difficulties of which the Revenue now complain stem, in part, from the way they argued the case.  They sought to identify complications and anomalies which might arise [37-39].  But they themselves had accepted that the paragraph 55 conditions had to be applied on a case by case basis by reference to the question whether the facts corresponded to those conditions. In those circumstances they must share responsibility for the conclusions which both Park J and the Court of Appeal reached both as to the date and the interpretation of the no possibilities test. If satisfaction of the paragraph 55 conditions does depend upon an assessment of the factual situation of the surrendering company, then it is easier to see why the courts should permit the availability of trans-national group relief to depend upon the facts at the date of claim as opposed to the facts at the date when the losses crystallised.

Park J [23(ii)] and the Court of Appeal [13] in Halsey recorded the Revenue’s submission at that time:

“Legislation of a Member State which imposes a blanket prohibition on intra-Community cross border surrenders of losses is not contrary to Community law, but on a case by case basis, may not be applied to any case the facts of which correspond to the circumstances described in paragraph 55 of the ECJ judgment” (my emphasis).

In accepting that submission, Park J interpreted paragraph 55 as meaning that the rule of UK tax law is valid despite the absence of a statutory exception:

“for losses which have not been used in the other Member State where the surrendering company is resident and are no longer capable of being so used.  The qualification to that ruling of the ECJ (the caveat, in Advocate General Geelhoed’s expression) is that, if on the facts of an individual case it can be shown that the losses have not been and cannot be used in the other Member State, the UK rule has to be disapplied to that particular case.” (my emphasis)

Similarly, when considering the nature of the ‘no possibilities’ test, Chadwick LJ took the view that objective facts were relevant both to past and future periods [49].

On appeal no-one sought to argue that that was wrong. The argument focussed on the date of the relevant objective facts and the meaning of the “no possibilities” test.  Since the Revenue had themselves argued for a case by case decision which tests the facts against the conditions identified in paragraph 55, it is not surprising that Park J should refer to the facts of an individual case and should have explained that when he referred to possibilities available he meant “recognised possibilities legally available given the objective facts of the company’s situation at the relevant time [33]”. The Court of Appeal concluded that there was no material difference between Park J’s interpretation and, as M&S contended, “the actual possibility of use in the light of both the taxpayer’s situation and the theoretical possibilities available under the local law” [46].            

The Revenue seem to me now to have adopted a different approach to the identification of the circumstances in which domestic law may or may not restrict trans-national group relief. They rely on principle and not on a factual, case by case basis for distinguishing between the circumstances in which prevention of trans-national surrender is lawful and those in which it is unlawful. Once again it is necessary to recall that the conditions in paragraph 55 are designed to prevent disruption to national systems by excluding the possibility of exercising a choice as to where the losses may be relieved, what the Netherlands would call 'trafficking in losses’.  The Revenue submits that once a surrendering company is permitted to satisfy the paragraph 55 conditions by reference to its particular factual situation at the time of its claim, it affords an opportunity for the surrendering company and its group to make a choice. It may make a choice which will trigger the opportunity to transfer the losses cross-border.  That choice may be made at any time up to the time of the claim.  Whether made for tax or commercial reasons, any choice which affects the availability of trans-national group relief jeopardises the balanced allocation of the power to impose taxes in the Member States concerned.

This appeal, so the Revenue contend, illustrates the exercise of a choice by the M&S surrendering companies which triggered the availability of  group relief in the UK. At the time their losses crystallised both surrendering companies were trading, but in the time between the end of the relevant accounting periods in which the losses accrued and the time of the claims on which they now rely (assuming they are valid claims), the decision of the Court of Appeal allowed them to choose to go into liquidation and thus, according to their case, to reap the benefit of group relief.  That choice, so the Revenue contend, was as much “an option to have their losses taken into account” ([46] M&S]) in Belgium or Germany or in the UK, as the choice of a trading company to traffic its losses to a Member State whose tax rates enabled those losses to be used to maximum effect.  There is, so the argument runs, no difference in principle between exercising a choice as to which country in which to group relieve losses and exercising a choice as to how to bring about a state of affairs which will afford an opportunity for carrying losses across a national border.

Accordingly, the principled objection to allowing the question whether the paragraph 55 conditions are satisfied to be answered by reference to the facts as they are at the time of the claim is that it gives an option or choice as to where the losses may be relieved. That option is recognised by the ECJ as substantially jeopardising fiscal sovereignty.

The implication of Park J's reasoning is that the only limit on the time by which the facts must satisfy the paragraph 55 conditions are limits imposed in domestic law on the time for making claims (6 years three months in the ‘pay and file’ years, and only when the Revenue choose to close the enquiries in the years thereafter).  This seems to me to take no account of the fact that it may well be within the power of the group to control events from the end of the relevant accounting period up to the time when it chooses to make its claim.  Park J acknowledged that the objection to permitting a claimant to rely on facts at the date of the hearing of an appeal is that it would permit the claimant to spin out time in the hope that facts would change. But he failed to acknowledge that the reason why that is objectionable is that it would afford an opportunity to bring about a situation in which the paragraph 55 conditions would be satisfied. In the period up to the appeal the claimant would be free to choose whether to bring about a situation in which losses could be transferred cross-border.

That objection applies with equal force to the time between the accounting period in which the losses crystallised and the time for making a claim. Freedom to choose in that period is no less objectionable. It amounts to a choice as to whether to transfer losses cross-border or not.

Chadwick LJ failed to recognise that the rationale for the paragraph 55 conditions did have relevance to the relevant date [36]. The option to have losses taken into account in a state in which a surrendering company is not resident does, in the ECJ's eyes, jeopardise fiscal sovereignty (ECJ) [46]. That option is available at any time between the end of the accounting period in which the losses crystallised and the date of the claim, on Chadwick LJ’s analysis.

The problem, however, remains as to whether it is open to this court to accept that principled objection.  The Revenue contended that it was open to this court to depart from the decision in the Court of Appeal because subsequent decisions of the ECJ demonstrated that the Court fell into error and we should follow those subsequent decisions.

I am more than happy, in the circumstances of the instant appeal, to follow the approach of Chadwick LJ in Condé Nast Publications Ltd v Customs and Excise [2006] EWCA Civ 976. He recognised that this court may refuse to follow its own earlier decision, by virtue of s.3(1) of the European Communities Act 1972 where:

“the judgment of the Court of Justice under consideration by this court in the earlier case had been the subject of further consideration - and consequent interpretation, explanation or qualification - by the Court in a later judgment” [44]. 

The ECJ has had opportunity to review what it said in M&S.  But in my judgment, contrary to the Revenue’s suggestion, it was not an opportunity which the court took.  It merely adopted its previous approach and reasoning in different circumstances; this does not permit this court to differ from the earlier decision of this court.

Oy AA [2008] STC 991 offers a paradigm of the circumstances in which a company exercised a choice which jeopardised fiscal sovereignty. Finnish law limited the right to make intra-group transfers, whereby taxable profit from one group company may be used to offset a loss made by another group company, to intra-group transfers between Finnish share companies.  A Finnish parent wished, for non-fiscal and genuine commercial reasons, to support an ailing United Kingdom established subsidiary by transferring profits in order to secure its financial position.  There was, plainly, no question of the double use of losses.  But the ECJ endorsed the Finnish prohibition on the grounds that the proposal constituted an election as to the state where the group’s profits should be taxed.  Legislation which precluded such a choice was proportionate to the objective of safeguarding the balanced allocation of the power to impose taxes between Member States, even though the transfer did not involve artificial arrangements designed to avoid tax. [63] and [64].

The importance of this case is the emphasis the Court placed on the freedom of the group to choose where its profits should be taxed, based on the principles it had expressed in M&S v Halsey.  It was the potential effect of that freedom which justified the Finnish prohibition against trans-national transfer of taxable profits. The Court regarded that feature as a sufficient and proportionate justification, even though there was no risk of using those transferred taxable profits more than once. [60].

The Court repeated that the fact that income of a parent cannot be taxed in the state where the subsidiary is resident is not a justification for systematic refusal to permit the cross-border transfer of losses. [53]  But:

“55.

The court has thus held that to give companies the right to elect to have their losses taken into account in the Member State in which they are established or in another Member State would seriously undermine a balanced allocation of the power to impose taxes between the Member States (see the Marks & Spencer case [46] and the Rewe Zentralfinanz case [42]”

It is true, as Mr Milne QC pointed out, that in a case such as Oy AA, where a group proposes to transfer profits as opposed to losses to a subsidiary resident in another Member State, it will always have a choice as to whether to transfer those profits.  Only where it is proposed to carry losses cross-border may the question arise as to whether the group had a choice. But the case serves to underline the importance of the issue whether it is open to a group to exercise a freedom to choose where its profits are taxed.

Although, that issue is more clearly highlighted in Oy AA, there is no hint that the Court purported to do more than to apply principles it had previously expressed in M&S.  There is nothing in that case which would justify a departure from that which this court decided in M&S.

Nor is such justification to be found in Lidl Belgium GmbH Case C-414/06 [2008] ECR 1-3601.  Lidl Belgium was resident in Germany with a permanent establishment in Luxembourg and sought to transfer losses from the permanent establishment to offset against German profits, contrary to German law. By virtue of a Double Taxation Convention, the profits of the permanent establishment would be taxed in Luxembourg.

Following Oy AA, the Court again emphasised the danger of giving companies the right to elect in which Member State  to have their losses taken into  account:

“31.

As regards the first of these justifications (the need to preserve the allocation of the power to impose taxes between the Member States), it should be noted that the preservation of the allocation of the power to impose taxes between Member States may make it necessary to apply to the economic activities of companies established in one of those States only the tax rules of that State in respect of both profits and losses (M&S paragraph 45 and Oy AA paragraph 54)

32.

To give companies the right to elect to have their losses taken into account in the Member State in which they are established or in another Member State would seriously undermine a balanced allocation of the power to impose taxes between the Member States, since the tax base would be increased in the first State and reduced in the second by the amount of the losses surrendered (see M&S paragraph 46, Oy AA paragraph 55)”

33.

With respect to the relevance of the first of these justifications in the light of the facts in the main proceedings, it should be pointed out that the Member State in which the registered office of the company to which the permanent establishment belongs is situated would, in the absence of a double convention, have the right to tax the profits generated by such an entity. Consequently, the objective of preserving the allocation of the power to impose taxes between the two Member States concerned, which is reflected in the provisions of the Convention, is capable of justifying the tax regime at issue in these proceedings, since it safeguards symmetry between the right to tax profits and the right to deduct losses.

34.

In circumstances such as the main proceedings, to accept that the losses of a non-resident permanent establishment might be deducted from the taxable income of the principal company would result in allowing that company to choose freely between the Member State in which those losses could be deducted (see to that effect Oy AA, paragraph 56)."

The Court also referred to the danger of the losses being taken into account twice [36]. But it acknowledged, following Oy AA, that national tax legislation precluding trans-national transfer of losses could be justified on the basis of the need to protect what the court described as symmetry between the right to tax profits and deduct losses.  The case is of importance in recognising that the danger to fiscal sovereignty caused by giving a group the right to elect was a sufficient and proportionate justification for the German tax regime [52] and [53].

But  there is, again, no hint that the Court thought it was departing from or going beyond what it had previously decided.  It had had every opportunity to do so.  Two years before Lidl, and a year before Oy AA, there had been a clear attempt to restrict the circumstances in which a law prohibiting trans-national transfer of losses could be said to be contrary to community law.  In Test Claimants in Class IV [2006] ECR 1-11673 Advocate General Geelhoed had described the article 55 conditions as a “caveat” to be “applied extremely restrictively”:

“65….(the caveat) should be applied extremely restrictively. It functions asymmetrically by, on the one hand, offering relief in cases where applying a source State's tax regulations results in losses for subsidiaries while, on the other hand, leaving extraordinary profits earned by subsidiaries operating in a more advantageous tax regime untaxed in the home State. The final result may be that, by virtue of this caveat, the Court has introduced an additional disparity in the interrelation between national tax systems, thereby further distorting the exercise of the freedom of establishment and free movement of capital within the Community. To put it somewhat differently, I see no reason why companies which decide to relocate their activities to another Member State, in full knowledge of the local tax legislation, should be awarded highly selective and distortional tax relief in the home State in circumstance where their source State activity incurs losses that cannot be offset in the latter State”.

But despite that simple and compelling plea for symmetry,  the Court has not, since then, taken the opportunity to refuse what the Advocate General described as highly selective and distortional tax relief.  It has never regarded the fact that a group has chosen to locate some of its activities in another Member State as, by itself, sufficient to justify the refusal to allow relief of trans-national losses.

I conclude that this court is bound by its previous decision in M&S v Halsey.  It is not open to this court to depart from the court’s previous decision that the question whether the paragraph 55 conditions are satisfied is to be answered by reference to the facts as at the date of claim ([36] of the Court of Appeal’s judgment).  Nor is it open to this court to depart from the previous court’s decision that the question for the court, under the second condition in paragraph 55, is whether there is, having regard to the objective facts at the time of the claim, a real, as opposed to a fanciful, possibility for losses to be taken into account in future periods.

I have stressed that it seems to me that the origin of this principle lies in the way the Revenue argued the case at that time. I have contrasted that approach with the approach now adopted by the Revenue because it underlies the authority of the principle expressed by the Court of Appeal. The Revenue cannot deflect the force of precedent by alighting upon an argument so different from that which they had previously adopted before this court.

I conclude, in answer to the first issue, that the Court of Appeal's decision as to the date for assessment of the paragraph 55 conditions is binding on this court. I would refuse the Revenue’s appeal on this issue.

It seems to me that that conclusion dictates the answer to the next issue, whether M&S can make sequential claims for the same loss of the same surrendering company in respect of the same period. It is true that neither Park J nor Chadwick LJ had to consider whether M &S was entitled to make sequential claims; there were no such claims before those courts. The Revenue's starting point is that the domestic statutory regime prohibits more than one claim in respect of losses in an accounting period from a surrendering company within the group. By Paragraph 73 of Schedule 18 to the Finance Act 1998:

“Withdrawal or amendment of claim

73.—

(1)

A claim for group relief may be withdrawn by the claimant company only by amending its company tax return.

(2)

A claim for group relief may not be amended, but must be withdrawn and replaced by another claim.”

This Regulation forms part of a statutory scheme designed to enable the Revenue to ensure that there is no mismatch between the amount claimed and the amount the surrendering company shows on its tax return. The scheme enables the Revenue to assess the validity of claims on the face of the documents that the statutory regime requires without being impeded by the likely complexities of the tax affairs of separate companies within the group. Thus a claim for group relief must be included in the claimant company's tax return for the accounting period for which the claim is made (paragraph 67(1)), and may be included in the original return or by amendment (paragraph 67(2)). The claim must specify the amount of relief quantified at the time of the claim and name the surrendering company (paragraph 68). Whilst a claim may be made for less than the amount available for surrender,

“A claim is ineffective if the amount claimed exceeds the amount available for surrender at the time the claim is made.”(Paragraph 69(2))

The FTT took the view that that provision rendered the M&S claims made prior to liquidation invalid. It thought that if a claim exceeded the amount which subsequently proves to be available for surrender, the claim was a nullity. On that interpretation, there was no objection to subsequent claims made at a time when they met the paragraph 55 conditions, because all the earlier claims which had not satisfied those conditions were a nullity. As the Upper Tribunal recognised [69], the earlier claims were not ineffective within the meaning of Paragraph 69(2): a claim which exceeds the amount which is available is simply excessive but it is a valid claim. Paragraph 69(2) deals with the different situation where a claim is made for more than the surrendering company, in its written consent, shows to be available for surrender. If that occurs, then the claim is invalid and the claimant company must make a new claim.

That interpretation of the meaning of an ‘ineffective’ claim finds support in other provisions which identify the circumstances when the statute taints a claim as ineffective. A claim is ineffective without a written notice of consent to the Revenue (paragraph 70). Without the contents specified in the Schedule, the notice of consent is ineffective (paragraph 71). These provisions, like paragraph 73, form part of a scheme, which attempts to ensure that any alteration in a claim is itself compliant with the requirements of the scheme. Paragraph 73 ensures that there is only one claim in respect of a particular surrendering company at a time.

Mr Ewart QC, on behalf of the Revenue, contended that, even if Chadwick LJ’s decision as to timing is accepted, it does not follow that it is permissible to make subsequent claims for the same loss; the only relevant claim is the first claim. It is the facts as they are at the date of that claim which must satisfy the paragraph 55 conditions and that date cannot change by putting in successive claims. There is no principle in community law which permits the relevant date to be changed and postponed until such time as the conditions are satisfied.

Mr Ewart draws an important and, as I shall explain, correct distinction between affording a claimant the opportunity to create a right to cross-border losses and the opportunity to prove that such a right has arisen. The community law principle of effectiveness is concerned with vindicating a right and not creating it. Mr Ewart argues that to recognise successive claims to the same losses is merely to allow a claimant to create a right, a right which did not exist at the date of the original claims because at that time the paragraph 55 conditions were not satisfied.

Moreover, since M&S has not withdrawn its first claim it follows, so the Revenue contend, that its second, third and fourth group relief claims are invalid. The first claims have not been withdrawn. Even if they were, there remains only one claim and that is the claim in the return. It cannot be amended because there is nothing to amend. The claim remains the same. It is a claim to those losses of the surrendering company in the relevant accounting period “which are available for surrender and capable of being claimed as group relief under s.403” (the wording adopted by M&S in their claims following the decision of the Court of Appeal, see e.g. their letter dated 20 March 2007). All that has changed, after the first claim, are the circumstances which have occurred subsequently.

The problem with these arguments is that they are inconsistent with the conclusion in the First Appeal. The consequence of that decision is that a claimant company is entitled to wait before it makes any claim to group relief until the circumstances are most favourable for demonstrating that the paragraph 55 conditions are satisfied. The only restriction is the time limited for making a claim. The initial claim must be made within a year of the filing date of the claimant company's tax return (paragraph 74(1)(a)); once an enquiry is opened, a claim may be made at any time up to 30 days after the conclusion of the enquiry (74(1)(b)), amendment of the tax return (74(1)(c))or appeal against any amendment (74(1)(d)). In the self-assessment years, inquiries are still continuing, following the joint references which triggered these appeals and time has, therefore, still not expired.

M&S, which made its first claims at a time when the conditions were not satisfied, and when it could not have known whether the conditions could be satisfied since it could not know what those conditions were, can surely not be worse off than if it had made no claim at all. On the Court of Appeal's understanding of the ECJ's decision, it makes no sense to deprive M&S of the ability to claim cross-border losses merely because its claims were premature. If it should have waited until it could satisfy the paragraph 55 conditions, and it was still in time to make claims to cross-border losses, it is difficult to see why it should lose that opportunity because it made its claims too soon. If the Revenue are correct in their essential argument that the conditions must be applied at the time the losses crystallised then the problem does not arise; no advantage is to be gained by making successive claims. But once it is accepted, as the Court of Appeal accepted, that a claimant may wait between the end of the accounting period in which the losses crystallised and the expiry of the time for making a claim, there is no reason why a claimant should forfeit the right to make a claim merely because it makes the claim too soon. The Court of Appeal has recognised a right to claim based on facts which arise after the end of the accounting period, and before the expiry of the time for making a claim. Since there is no restriction against withdrawing a claim and advancing a new claim within that period, there is no good reason to prevent M&S doing so for the purpose of satisfying the paragraph 55 conditions. To refuse M&S the right to withdraw its earlier claims would put it at an unjustifiable disadvantage as against other potential claimants who have made no claim at all. If the only inhibition on waiting is the time limit for bringing claims, there can be no reason for refusing to allow M&S to withdraw such claims made at a time when the facts do not satisfy the paragraph 55 conditions, and rely on a claim made at a time when they do. The only time limit for such withdrawal is that which is consequent on the time limits within paragraph 74.

That result may be achieved, in compliance with paragraph 73, by M&S withdrawing the earlier claims and amending its return to make the claim at a time when the facts do satisfy the conditions in paragraph 55 pursuant to paragraph 75(6) of Schedule 18.

The mechanics for achieving that solution were explained with care and in full detail by the Upper Tribunal between paragraphs 67 and 113. The Upper Tribunal has demonstrated the flexibility of the domestic statutory scheme which achieves a result whereby:-

“in their final form, the tax returns of the claimant and surrendering companies accurately reflect amounts eventually shown to be available for surrender, as supported by corresponding notices of consent. Further, the processes and adjustments required to reach that final result may continue throughout the period during which it is open for a group company to make a group relief claim (which, in practice, under self-assessment, is a generous period).” [86]

As the Upper Tribunal recognised, paragraph 69(2) makes no sense in relation to a group relief claim to losses surrendered by a foreign subsidiary. In contrast to the tax return from a UK surrendering company, which will reveal the information necessary for the purposes of s.403(1), a foreign subsidiary will not submit any UK tax return and any equivalent return will not demonstrate the extent to which any losses which it is sought to surrender satisfy the paragraph 55 test. [UT 107]. Paragraph 69 should be ignored, as the Upper Tribunal proposed [112]. Any formal statutory requirement must be “disapplied or moulded” so that the right is not rendered “practically impossible or excessively difficult” (Lord Nicholls in Autologic plc v IRC [2006] 1 AC 118 [17] and [30].

The issue, however, is not one of mechanics but of principle. The Revenue's objection is that a claimant should not be permitted to delay making a claim until it can satisfy the paragraph 55 test and, accordingly should not be permitted to withdraw earlier claims, which do not satisfy that test. But, like the Upper Tribunal, I see no reason why it should not. Either the Schedule permits such a course or it must be moulded for that purpose. Once it is acknowledged, as the Court of Appeal decided, that a claim may be delayed from the accounting period in which the losses claimed crystallised to the end of the time for making a claim, there can be no reason not to permit a series of claims being made. It seems to me that the Revenue's objection can only succeed if they are correct in their essential argument that a claimant cannot rely upon any facts other than those which exist at the time when the losses claimed crystallised. Once it is accepted that facts which arise subsequently, and up to the expiry of the period for making a claim, are relevant, the objection becomes a mere question of machinery.

I conclude that the decision of the Court of Appeal dictates that the claimant M&S is permitted to make successive claims to the same loss and rely on the claim which satisfies the paragraph 55 criteria, and then withdraw any earlier claims to the same surrendered losses. I would refuse the Revenue’s appeal on this issue.

I have accepted that the Revenue are correct in their submission that the principle of effectiveness is concerned with vindicating not creating a right. That proposition is relevant to this issue. In the years ending 31 March 1998 and 31 March 1999, MSG suffered over £21 million losses. The latest date on which M&S could have claimed group relief for the last pay and file accounting period (31 March 1999) was 30 June 2005. The latest possible date on which any taxpayer could have claimed such relief, for accounting periods ending 30 June 1999, was 30 September 2005. The ECJ released its judgment, enunciating for the first time the paragraph 55 conditions, on 13 December 2005. In those circumstances, M&S contend that it should have a reasonable time from the date of the judgment to put itself in a position in which it can satisfy the ‘no possibilities’ test. The principle of effectiveness demands that there should be a reasonable period in which to demonstrate that it satisfies the paragraph 55 conditions, in particular because claims to those losses were made in time, the Revenue were well aware of those claims and, accordingly, in the particular circumstances of these claims, the principle of legal certainty, which, as established jurisprudence acknowledges, usually trumps the principle of legal effectiveness, can have no application. Insistence on the time limit rendered the exercise of the right to claim cross-border losses virtually impossible or excessively difficult.

The Upper Tribunal reversed the decision of the FTT. The FTT concluded that the principle of effectiveness was breached since no taxpayer could have foreseen the ECJ's identification of the paragraph 55 conditions, and thus the exercise of the right to claim such losses was impossible or excessively difficult. The case for applying the principle of legal certainty was, so it said, ‘much weaker’, because M&S had made claims within time and was merely purporting to 're-exercise' its right to claim such losses.[49].

The Upper Tribunal concluded that the principle of effectiveness has no application when a claimant seeks an opportunity to bring about a state of affairs which would create new rights which the claimant did not already have. At the time the claims were made, M&S did not satisfy the ‘no possibilities test’ and thus could not have satisfied the paragraph 55 conditions within the time limits provided in domestic law. Absent such right, the principle of effectiveness could not be invoked to create a new right subsequently. [158]

This disagreement encapsulates the dispute before this court. The relevant jurisprudence establishes that a Member State may impose a reasonable time limit in the interests of legal certainty (Aprile, Case C-228/96, [1998] ECR 1-7141[19] and Fleming v HMRC [2008] 1 WLR 195 [79(a)]). But such a time-limit must not render virtually impossible or excessively difficult the exercise of rights conferred by Community law: Aprile [19]. Both the FTT and the UT discussed the line of cases concerned with the reduction of a time limit which has the effect of taking a right away without adequate transitional arrangements (discussed, for example, in Case C-62/00 Marks & Spencer plc (M&S I) [2003] QB 886). I cannot see how that line of cases has relevance to these claims. The time limit of 6 years and 3 months was in place, M&S’s claims were made within that period and were found not to have satisfied the paragraph 55 conditions.

There can be no doubt but that M&S could not have foreseen the ECJ's identification of those circumstances in which a Member State's prohibition against the use of cross-border losses would offend Article 43. Whilst the origin of the paragraph 55 conditions lay in the Advocate General's opinion, none of the parties had themselves advanced or debated those conditions. But non sequitur that M&S’s rights were rendered virtually impossible or excessively difficult to exercise. Such a complaint must be founded upon the existence of community rights at the time when the impugned time limit or retrospective imposition of such a limit serves to forestall their exercise. The logically prior question is whether, at the expiry of the time limit for making a claim, M&S had a right to claim the MSG losses.

At the time M&S made its claim to the losses sustained by MSG, it had no community law right to make such a claim. The prohibition against such a claim was lawful because M&S did not satisfy the conditions identified by the ECJ in paragraph 55. The ECJ has espoused the principle that, provided that the time limits are not discriminatory and do not render the exercise of Community law rights virtually impossible or excessively difficult in practice, a Member State may lay down reasonable time limits even if their effects to deprive a claimant of such a right (Haahr Petroleum, Abenra Havn and Others [1997] ECR 1-4085 [48]). That case concerned, like Aprile and M&S I, the propriety of a time limit for claims to repayment. There is no principle that a reasonable time must be afforded to a claimant in which to bring about the circumstances which would generate the community law right. The error of the FTT lay in the assumption that M&S had a right at the time it made its claim; on the findings of fact, at that time it had no such right and the principle of effectiveness cannot be invoked to create one. In my view the Upper Tribunal was correct and the ‘pay and file’ claim in respect of MSG is time-barred. I would uphold the decision of the Upper Tribunal.

The piquancy of that conclusion lies in the contrast with my conclusion in relation to successive claims. But for the decision in the First Appeal, the principle which dictated the Upper Tribunal’s decision and mine on this point, that effectiveness is concerned with vindicating not creating a right, would have prevented successive claims. It is not community law which sanctions successive claims, but the decision in the First Appeal, which, in my view, is inconsistent with a prohibition of such claims.

The Revenue seek permission to appeal against the Upper Tribunal’s decision that the ‘no possibilities’ test is to be applied on a euro by euro basis so that at the time the claim is made the paragraph 55 criteria may be fulfilled even if a proportion of the losses do not satisfy them. Adopting that approach, the Upper Tribunal concluded that once liquidation had commenced, at the time of the second and subsequent claims, the no possibilities test was satisfied “to the extent that losses could not be utilised against income arising in the normal course of the liquidation process.” [164(3)]

There was a dispute between the Tribunal and the Revenue, which arose when the Revenue sought permission to appeal as to what the Upper Tribunal decided. It is plain to me that the issue needs to be resolved. The Upper Tribunal acknowledged that their decision involves assessment on a euro by euro basis [48-49] [168]. I would grant permission to appeal this point.

The facts, as found by the FTT, demonstrate the consequences feared by M&S should the Revenue's argument, that the existence of some possibility of use in relation to some part of the loss excludes a claim in respect of any part of the loss, succeed. Both MSG and MSB had ceased trading by 31 December 2001 and had not resumed in the period of almost 5 years before they were put into liquidation in October 2006. Under German law the objects of a company in liquidation are restricted to completion of the liquidation, and under Belgian law a company in liquidation only exists for the purposes of that liquidation and no new activities or investment may take place. The only reason that the Belgian and German subsidiaries were retained between 2002 and 2006 was to meet an unforeseen requirement of some formality in relation to the claims such as the execution of documents (according to the evidence of the group company secretary, Oakley (1st witness statement, 1 March 2006). Whilst in liquidation MSG earned a profit of €177,000 attributable to interest from the sale of its assets. It would take over 700 years to use up its €125,000,000 tax losses against profits earned at that rate. If the Revenue are correct, the existence of the possibility of earning so paltry a sum excludes the use of all its losses.

It seems to me that the problem of precluding the cross-border use of losses merely because one euro of loss may be used in another Member State relates to the essential primary issue between the Revenue and the taxpayer. It is not a distinct question. If the Revenue are correct, the problem does not arise. At the end of the accounting period in which the losses were sustained, the paragraph 55 criteria are applied by asking whether past losses could have been set off and whether future losses could be used. Such an all or nothing approach does not admit of any question as the likelihood of a proportion of losses being used in the future because it looks at the characteristics of the losses in question and merely applies the paragraph 55 criteria to losses of the type or quality identified under the domestic law in question.

But in obedience to the previous decision of the Court of Appeal, it seems to me that the conclusion of both of the Tribunals necessarily follows. Once it is accepted that the assessment of the paragraph 55 criteria involves a factual question based on the factual position of the subsidiaries which suffered the losses, it seems to me that the absurdity of excluding all losses simply because one euro of loss may be used should be avoided. I agree with the conclusion of the Upper Tribunal at [49] and would dismiss the Revenue’s appeal on this issue.

The final issue concerns the method by which foreign losses are afforded group relief within the United Kingdom. Although the controversy is complex, it involves an important question of principle. Permission was refused to both sides, although M&S is content with the conclusion adopted by both the FTT and the Upper Tribunal, which became known as 'Method E'. It is tempting to resort to the principle that the matter is best left to the distinguished and experienced tribunals who have already decided in favour of Method E. I do not believe that it is open to this court to invoke that principle in a case where the problem is novel and finds no analogy in the type of problem with which tax experts and tribunals have, in the past, been faced. I am of the view that this court must reach its own conclusion, tutored, of course, by those much respected members of the two tribunals who have grappled with this problem previously. I would grant permission to both sides.

The problem arises out of the need to ensure that the losses of a subsidiary in one Member State are not subject to group relief to a greater extent than that which would be available had that subsidiary been resident in the Member State in which the relief is claimed. It is convenient to consider the losses suffered by MSG. All agree that the unused German subsidiary's losses cannot be used for relief to a greater extent than would be allowed had the subsidiary been resident in the UK and subject to UK tax rules. For that purpose, UK tax rules must be applied to the German unused losses and that application sets a cap on the amount of German losses which may be used for relief in the UK.

No problem arises where one category of loss is recognised under German tax law but not under UK tax law. It is agreed that, if any of the German losses claimed could not be claimed under UK law, they cannot form part of the group relief claim in the UK. But the problem does arise in relation to differences in timing. German tax law recognises losses for tax purposes in accounting periods which are different from those periods in which they would be recognised under UK tax laws. For example, capital expenditure in some tax systems is deductible in the same way as depreciation of a capital asset - over the life time of the asset, but in the UK it is deductible under a system of capital allowances, which accelerates the deduction to a greater extent than the actual depreciation for accounting purposes. Differences in timing also occur in relation to rental income. The application of Method E adopted by both the FTT and the UT, with which M&S is now content, applies not only UK tax law in relation to categories of loss sustained by the German and Belgian subsidiaries but also applies UK tax law so as to identify the periods in which those losses were sustained. UK tax law may throw losses sustained in Germany into a different accounting period under UK timing rules. For example, a loss made under German rules in year 2 may be recognised and afforded relief under UK rules in year 1. In such circumstances M&S contend, and the Tribunals agreed, that it may be afforded group relief in year 1, even though under German tax rules, the German subsidiary suffered no loss in year 1.

The Revenue contend that no such result is possible. No system of quantification can be permitted which allows a loss to be claimed in a period in which, in Germany no loss was sustained. For example, in 2002 no loss was sustained in Germany. But the effect of the application of Method E is to permit losses to be relieved in 2002 in the UK. No principle of EU law requires the German losses to be relieved to a greater extent than would be the case if they were claimed in Germany.

To resolve this difficulty it is necessary to analyse Method E. Under Method E the process is as follows:

The dispute arises because at stage (iv) the effect of reaching an equivalent United Kingdom tax loss is that it may be suffered in an accounting period which differs from that in which it was in fact sustained in Germany. The equivalent United Kingdom amount of loss is not necessarily attributed to the same accounting period or not only to the same accounting period as in Germany.

The Revenue contend that a method of quantification should be adopted which avoids a result whereby losses may be afforded group relief in an accounting period in which in Germany no unused losses were in fact sustained and in which in fact there was a profit. Under the Revenue's method, called Method F, losses minus those utilised are calculated independently under German tax rules and independently under United Kingdom tax rules and group relief is afforded to the lower of the two taxable results.

The difference between the results given by adopting Method E rather than method F is substantial. For example, in 1999, it leads to a difference of just over £2 million.

The principle to be applied to resolve the controversy is to be identified, the Revenue suggest, in the opinion of Advocate General Geelhoed in Test Claimants in Class IV of the ACT Group Litigation Case C-374/04 [2006] ECR 1-11673. The Advocate General recognised that disparities in tax treatment are inevitable where discrete national tax systems co-exist within the EU [43]. Distortions resulting from such disparities do not fall within the scope of the free movement provisions of the Treaty [46]. But obstacles which occur as a result of the rules of just one tax jurisdiction are discriminatory and may offend Article 43 [46].

Those principles are echoed by the ECJ which acknowledges that disadvantages which are the result of "the parallel exercise of fiscal sovereignty by the Member States" are outwith the scope of Article 43 (e.g. Orange European Smallcap Fund Case C-194/06 [2008] ECR 1-03747 [37]).

The Revenue contend that under the UK tax regime, group relief is afforded on an accounting period by accounting period basis. It is fundamental that, if UK tax law would recognise a loss in an accounting period in which there was in fact a German profit, a prohibition against claiming group relief in that accounting period for losses sustained in Germany is compatible with community law. The fact that application of UK tax law would place a loss sustained by MSG in an accounting period in which no such loss was sustained under German tax law does not entitle the loss to be set against the profits of M&S in the UK in that accounting period. That is merely the inevitable consequence of the two different tax systems. The restriction on such a claim is not incompatible with Article 43.

I disagree. M&S seeks to set against its UK profits losses sustained by its subsidiary in Germany, as if those losses were sustained by a subsidiary resident in the UK. It claims no more and no less. If the losses had been sustained in the UK, it seems to me that there would be no question of timing differences leading to the loss of relief in respect of a proportion of unutilised losses. The effect of the application of UK tax rules may be to shift losses sustained in Year 2 under German tax rules into Year 1, if the subsidiary had been resident in the UK. Those losses should be afforded relief in Year 1 under UK rules. It is nothing to the point that that would not be the appropriate year under German tax rules. The effect of the application of UK tax rules is to convert the German losses into losses sustained in year 1 to be set against UK profits in the same accounting period, i.e. year 1. That is not to cut across UK tax principles but to apply them.

The consequence of the Revenue's method is to deprive M&S of relief for losses sustained in Germany in circumstances where it would not be refused relief had those losses been sustained in the UK. Method E does not give the parent greater relief than would have been available had its subsidiary been resident in the same state as the parent, whether in Germany or in the UK. It does not seem to me to matter that the losses are allowed in different accounting periods from those in which they would be allowed in Germany. No relief is to be afforded to losses which would not be relieved in the UK. As the FTT put it:

“Once you move from identifying the local losses (computed under local rules) to identifying their equivalent under UK rules, you also have to move from local time of recognition to UK timing of recognition” [7].

Method E does not result in a group relief claim for an amount more than could be claimed were the subsidiary to have been resident in the UK. The re-allocation of losses to a different period in the UK is merely the result of the application of UK tax law. I would dismiss the Revenue's appeal on this point.

I would grant permission in relation to all those issues in respect of which permission is needed and refuse the appeals of the Revenue in relation to issues (i) (ii) (iii) and (v) and of M&S in relation to issue (iv).

I agree.

I also agree.