Loss Relief Group Litigation Order, Claimants Listed In v Revenue & Customs

The basic question which I have to decide on this application by the claimants in the Loss Relief Group Litigation Order (“the Loss Relief GLO”) is whether the High Court should at this stage in the group litigation make a reference to the Court of Justice of the European Union (“the CJEU”, formerly the European Court of Justice (“the ECJ”)) for a preliminary ruling under Article 267 TFEU on the question whether claims for cross-border group relief could in principle be made by UK-resident claimant companies at the relevant times in various forms of corporate group structure which differ from the simple structure (surrender of losses by an EU or EEA resident subsidiary to its UK resident parent company) which was considered by the Grand Chamber of the ECJ in its seminal decision in Case C-446/03 Marks & Spencer PLC v Halsey (Inspector of Taxes) [2005] ECR I-10837, [2006] Ch 184, [2006] STC 237 (“M&S v Halsey”).

In M&S v Halsey the UK resident parent company of the group, Marks & Spencer PLC, sought to offset losses incurred from the later 1990s to 2001 by its subsidiaries in France, Germany and Belgium against its UK profits, by way of claims for group relief from corporation tax under section 402 of the Income and Corporation Taxes Act 1988 (“ICTA 1988”). Under the relevant UK statutory provisions, the claims to group relief could not succeed, and were accordingly rejected by the Revenue, because only losses of a UK resident company (or, after 2000, losses of a non-resident company that carried on a trade through a UK branch) could be surrendered by way of group relief. It was this territorial restriction on the scope of group relief which M&S challenged as infringing the rights to freedom of establishment and free movement of capital then contained in Articles 43 and 56 of the EC Treaty, and now contained in Articles 49 and 63 TFEU.

The challenge enjoyed a limited measure of success, although quite how limited remains a subject of acute controversy. Before describing the main areas of disagreement about the effect of the ECJ’s ruling, I will first summarise what the ECJ actually decided. First, it held that the territorial restriction of group relief to UK companies constituted a restriction on the right of establishment of a UK resident parent company such as M&S (paragraphs 27 to 34 of the judgment of the Court). Next, the ECJ considered the factors relied upon by the UK and other member states which had submitted observations as justifying the restriction. Those factors were, in brief, (a) the need to protect a balanced allocation of taxing power between the member states concerned, with profits and losses being treated symmetrically; (b) the risk of losses being taken into account twice, if they were relievable in the parent company’s member state; and (c) the risk of tax avoidance if the losses were not taken into account in the subsidiary’s state of establishment (see paragraph 43). The Court discussed and gave its endorsement to each of these factors in paragraphs 44 to 49, before concluding in paragraph 51:

“In the light of those three justifications, taken together, it must be observed that restrictive provisions such as those at issue in the main proceedings pursue legitimate objectives which are compatible with the Treaty and constitute overriding reasons in the public interest and that they are apt to ensure the attainment of those objectives.”

Finally, the Court considered whether the restriction “goes beyond what is necessary to attain the objectives pursued” (paragraph 53), or in other words the question of proportionality. The Court’s reasoning and main conclusion on this issue were as follows:

“54.

Marks & Spencer and the Commission contended that measures less restrictive than a general exclusion from group relief might be envisaged. By way of example, they referred to the possibility of making relief conditional upon the foreign subsidiary’s having taken full advantage of the possibilities available in its member state of residence of having the losses taken into account. They also referred to the possibility that group relief might be made conditional on the subsequent profits of the non-resident subsidiary being incorporated in the taxable profits of the company which benefited from group relief up to an amount equal to the losses previously set off.

55.

In that regard, the Court considers that the restrictive measure at issue in the main proceedings goes beyond what is necessary to attain the essential part of the objectives pursued where:

- the non-resident subsidiary has exhausted the possibilities available in its state of residence of having the losses taken into account for the accounting period concerned by the claim for relief and also for previous accounting periods, if necessary by transferring those losses to a third party or by offsetting the losses against the profits made by the subsidiary in previous periods, and

- there is no possibility for the foreign subsidiary’s losses to be taken into account in its state of residence for future periods either by the subsidiary itself or by a third party, in particular where the subsidiary has been sold to that third party.

56.

Where, in one member state, the resident parent company demonstrates to the tax authorities that those conditions are fulfilled, it is contrary to arts 43 EC and 48 EC to preclude the possibility for the parent company to deduct from its taxable profits in that member state the losses incurred by its non-resident subsidiary.”

It was accordingly only at this final stage in the analysis that the challenge to the UK’s group relief regime succeeded, and then only to the extent of the cumulative test enunciated in paragraph 55. This test has come to be known as the “no possibilities” test. It was repeated by the ECJ, in materially identical words, in paragraph 59, and again in the dispositif at the end of the judgment.

The first main area of dispute generated by the judgment in M&S v Halsey concerns the interpretation of the no possibilities test: what exactly does it mean, and at what date does it have to be applied? The main stages in the evolution and resolution of this dispute, to date, have been as follows.

The reference to the ECJ was made by Park J, on the hearing of an appeal by M&S from the Special Commissioners who had upheld the Revenue’s refusal of the claims to group relief. When the case returned to Park J after the ECJ had given its judgment on the reference, he heard argument on and decided a number of questions of principle before remitting the matter to the Special Commissioners for them to make further findings of fact and finally determine the appeal: see Marks & Spencer PLC v Halsey (Inspector of Taxes) [2006] EWHC 811 (Ch), [2006] STC 1235 (“M&S (Chancery)”).

In relation to the no possibilities test, Park J held that when the ECJ referred to “possibilities available” it meant “recognised possibilities legally available given the objective facts of the company’s situation at the relevant time”: see paragraph [33]. He then gave some helpful examples at paragraphs [37] to [39]. As to the relevant time, he considered that there were three possibilities: the end of the accounting period of the subsidiary in which the loss was made; the time or times when M&S made the claim or claims for group relief; and the time when the appeal on the question was decided by the Special Commissioners (paragraph [43]). Park J came down in favour of the second possibility, holding that the first was “too soon”, because it would probably rule out group relief in every case, while the third would allow the parent company to “spin out time before the matter came to appeal in the hope that by then the facts would have changed and the appeal would succeed” (paragraphs [44] and [45]). By contrast (see paragraph [46]):

“… time (2) in my view provides a rational basis for applying para 55. If a company claims group relief at a time when the para 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.”

The reasoning and conclusions of Park J on the no possibilities test were substantially upheld by the Court of Appeal: see Marks & Spencer PLC v Halsey (Inspector of Taxes) [2007] EWCA Civ 117, [2008] STC 526 (“M&S (CA) I”). The leading judgment was given by Chadwick LJ, with whom Tuckey and Jacob LJJ agreed. On the timing issue, Chadwick LJ relied on the fact that the question of the compatibility of the domestic group relief regime with EU law “does not arise until a claim for group relief is made by the claimant company” (paragraph [36]). As to the type of possibility which the test involved, he considered that “no possibility” meant “no real possibility”, in the sense of one which could not be dismissed as fanciful; and that the test was not to be equated with a test of “little or no likelihood”, because “a possibility may exist even where there is little or no real likelihood that the event will happen” (paragraph [49]).

The issue came before the Court of Appeal for a second time in 2011, on appeals and cross-appeals by the Revenue and M&S from the decision of the Upper Tribunal (Warren J (P) and Judge Sadler, [2010] UKUT 213 (TCC), [2010] STC 2470), which had allowed in part an appeal from the final determinations of the group relief claims by the First-tier Tribunal (Judge Avery-Jones and Judge Gammie QC): see Marks & Spencer PLC v Halsey (Inspector of Taxes) [2011] EWCA Civ 1156, [2012] STC 231 (where it is reported as Marks & Spencer PLC v Revenue & Customs Commissioners) (“M&S (CA) II”). The members of the Court of Appeal on this occasion were Lloyd, Moses and Etherton LJJ. The leading judgment was delivered by Moses LJ, with whom the other two members of the Court agreed.

One of the questions which the Court of Appeal had to consider was whether it was bound, in accordance with the usual rules of precedent, to follow its earlier decision on the no possibilities test in M&S (CA) I, or whether the earlier decision was no longer binding on the Court in the light of subsequent decisions of the ECJ. On this question, the Court held that it was bound by its previous decision: see the discussion of the subsequent European case law in paragraphs [35] to [45], which led Moses LJ to conclude at paragraph [46]:

“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 para 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 para 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.”

It is nevertheless clear that the Court of Appeal saw considerable attraction in the argument now advanced by the Revenue that the date of claim was the wrong date at which to decide whether the no possibilities test was satisfied, because it allowed the group to make a choice about the date of surrender of the losses at any time up to the time of the claim. The Revenue’s argument, which according to the Court was markedly different from the approach which they had adopted in M&S (CA) I, was that such a degree of choice “jeopardises the balanced allocation of the power to impose taxes in the member states concerned”: see paragraph [27]. Moses LJ went on in paragraphs [30] to [32] to criticise the reasoning of Park J and Chadwick LJ in taking the date of the domestic claim to group relief as the only date at which the no possibilities test had to be satisfied. As Moses LJ said in paragraph [30]:

“The implication of Park J’s reasoning is that the only limit on the time by which the facts must satisfy the para 55 conditions are limits imposed in domestic law on the time for making claims (six years three months in the “pay and file” years, and only when the Revenue chose 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.”

The present position is that the Revenue have been granted permission to appeal to the Supreme Court on various issues, including the no possibilities test, and the appeal will be heard by the Supreme Court in June 2013. It is, however, far from clear that the Supreme Court will be able to provide definitive guidance on the interpretation of the no possibilities test without making a further reference to the CJEU. On behalf of the claimants, Mr Aaronson QC went so far as to submit that the chances of the Supreme Court deciding the question without a further reference were “infinitesimal”. He pointed out in this context that the European Commission announced on 27 September 2012 that it had decided to refer the United Kingdom to the CJEU “for its tax legislation on cross-border loss relief”, as it is put in the press release issued by the Commission on that date. The basis of the proposed infraction proceedings is that the UK has failed properly to implement the previous decision of the ECJ in M & S v Halsey. The implementing legislation was contained in the Finance Act 2006, and reflected the Revenue’s views on the correct interpretation of the no possibilities test. If the Supreme Court does consider it necessary to make a further reference, it is unlikely that the CJEU would deliver its judgment before late 2015, and only then would the English courts be able to begin the process of interpreting and applying the Court’s answers to the questions referred. That process might well, in turn, involve a further round of domestic appeals, and it could easily be 2018, says Mr Aaronson, before the law on the no possibilities test is finally settled.

On behalf of the Revenue, Mr Ewart QC did not agree that the Supreme Court would be bound to make a further reference to the CJEU. He said the Revenue would not be arguing in the Supreme Court that the CJEU should be asked to reconsider the decision in M & S v Halsey, and that the Revenue’s submissions would be focused on the correct interpretation of the no possibilities test in the form in which it was enunciated by the ECJ. The Supreme Court will also have to consider various other issues, which Mr Ewart briefly explained to me, but again I did not understand it to be his view that any of them would be likely to necessitate a further reference.

In view of these conflicting submissions, I am unable to reach any firm conclusion about the likelihood of a further reference to the CJEU in connection with the no possibilities test. Further, I think it would in any event be unseemly for me to express a view on a question that only the Supreme Court can decide, namely whether it considers a further reference necessary in order to enable it to dispose of the forthcoming appeal. I am, however, satisfied, on the basis of the limited argument I have heard, that there is at least a significant possibility of a further reference being made, and if that were to happen the ultimate resolution of the no possibilities issue could well be delayed for at least a further two years beyond the date when the Supreme Court delivers its judgment, probably in the Autumn of this year.

I now turn to the other main area of dispute in the wake of M & S v Halsey, namely the structural questions to which I adverted at the start of this judgment. These questions arise at an earlier stage in the analysis than the no possibilities test, and go to the issue whether the freedom of establishment of a group company is relevantly engaged in the first place by a restriction which requires justification if it is not to infringe Article 49 TFEU. (It is now clear, in the light of subsequent European case law, that the right to free movement of capital under Article 63 is not engaged by a legislative scheme which is, broadly speaking, confined to corporate groups, so it is only freedom of establishment which is now in issue).

According to the claimants, there are three types of corporate structure in particular, apart from the basic one considered in M & S v Halsey itself, in respect of which it is necessary to know whether a claim for cross-border loss relief could validly be made in reliance on EU law at the relevant times. Those structures are:

In contrast with the no possibilities test, very little progress has been made in resolving the grouping issues, although everybody agrees that, unless they are somehow rendered academic, their resolution will require a reference to the CJEU. Instead, the issues have become bogged down in a procedural impasse before the First-tier Tribunal (“the FTT”). It would be tedious, and unnecessary, to relate the history of this procedural wrangle in full detail. I think it is enough to say that, after debating the matter inconclusively with the Revenue in correspondence, Dorsey & Whitney (Europe) LLP (the appointed lead solicitors in the Loss Relief GLO) sought to bring it to a head by making applications to the FTT for closure notices under paragraph 33 of Schedule 18 to the Finance Act 1998 (“FA 1998”) on behalf of three representative applicants. The applicants were Finnforest UK Ltd (“Finnforest”), Card Protection Plan LLP (“CPP”) and ExxonMobil Chemical Ltd (“ExxonMobil”), each of which had a different group structure. CPP’s structure was of type (a) in paragraph 17 above, while ExxonMobil’s structure was of the fourth type involving a non-EU/EEA parent. Finnforest’s structure was a further variant, involving a loss-making company in a non-UK member state surrendering losses to a profitable sister company in the UK, but with a parent in a non-UK member state: in other words, a structure of type (a) but with the common parent company resident in another member state. The applications sought a direction from the FTT that the Revenue issue closure notices within a specified time in respect of several open enquiries into the corporation tax returns of the applicants for the relevant years. The only open items in relation to the enquiries were the cross-border group loss relief claims.

At first sight, it may seem hard to understand why the Revenue had insisted on keeping the enquiries open, given their stance that (a) cross-border group relief was in principle available only in the simple vertical group structure considered by the ECJ in M & S v Halsey itself, and (b) the no possibilities test had to be satisfied at the end of the accounting period in which the loss was made by the surrendering subsidiary (with the practical consequence that the relief could virtually never be granted). If either or both of those contentions were correct, the claims could not succeed: so why should the claims not be refused and the enquiries closed, leaving the applicants to appeal, and thereby enabling the FTT to make a reference to the CJEU on the grouping issues? The Revenue’s response to this was that it was reasonable to keep the enquiries open until the full facts had been established, not only in relation to the group structures, which were often of considerable complexity, but also in relation to the application of the no possibilities test on the assumption that it had to be satisfied only when the claim for relief was made, which was of course what the Court of Appeal had held in FII (CA) I. To this the applicants replied that it would be burdensome and oppressive to require them to provide detailed evidence in relation to the no possibilities test, which involved factual issues of considerable complexity and the collection of a great deal of information, when those questions would never arise if the Revenue were right in either of their basic contentions.

With a view to obtaining a reference of the grouping issues to the CJEU, three possibilities were canvassed by the applicants in their submissions to the FTT. The first, and most straightforward, was that the FTT should direct the closure notices to be issued, so that the taxpayers could appeal and a reference could be made in the context of those appeals. The second was that the group structure issues should be referred to the FTT as a preliminary matter under the procedure contained in paragraph 31A of Schedule 18 to FA 1998, which provides as follows:

“31A(1) At any time when an enquiry is in progress into a company’s tax return any question arising in connection with the subject-matter of the enquiry may be referred to the tribunal for determination.

(2)

Notice of referral must be given –

(a)

jointly by the company and an officer of Revenue and Customs,

(b)

…

(c)

to the tribunal.”

Since such a notice has to be given jointly by the company and the Revenue, it follows that this procedure can only be invoked if both sides agree. The third possibility was that the group structure issues should be considered as a substantive preliminary issue of law in the context of the closure notice applications, thereby again enabling a reference to be made by the FTT. In Revenue and Customs Commissioners v Vodafone 2 [2006] EWCA Civ 1132, [2006] STC 1530, the Court of Appeal had held that paragraph 33 of Schedule 18 confers jurisdiction on the Special Commissioners (now the FTT) to decide incidental questions of law arising on an application for a closure notice, and that a reference could be made to the ECJ in the course of determining such a question of law.

The FTT (Judge Kempster and Ms Hunter) heard the applications on 25 November 2010, and gave their decision nearly six months later on 20 May 2011. They declined to direct the issue of closure notices, on the ground that it would be inappropriate to do so while the facts relating to the grouping issues and the no possibilities test were still under investigation. As they explained in paragraph [69] of the decision (see Finnforest UK v Revenue & Customs Commissioners [2011] UKFTT 342 (TC), [2011] SFTD 889):

“The taxpayers do not allege HMRC are being unreasonable in asking questions or expecting detailed answers – rather they would prefer, for good reasons articulately submitted, to have to cross that bridge only after the grouping test challenge has been resolved. While we sympathise with that desire and see some merits in it, to order the closure of the entire enquiries and expect HMRC to rely on pre-hearing disclosure on major information gathering is, we consider, not the correct action. Given the accepted volume of information still to be provided, it is not reasonable to stipulate a specified period within which HMRC should complete their enquiries. Accordingly, we would not make a para 33 direction to HMRC to close the enquiries.”

In relation to the applicants’ second proposal, namely a joint referral of the grouping issues to the FTT pursuant to paragraph 31A of Schedule 18, the FTT expressed the view that the possibility of such a referral was “the most constructive method of advancing this dispute”: see paragraph [71]. As they pointed out, enquiries in relation to the no possibilities test could continue while the referral on the grouping issues was before the FTT. They therefore said (ibid):

“We would urge the parties to devote some effort to attempting to agree the terms of such a referral. If agreement proves not possible then it is always open to the taxpayers to renew their closure notice applications.”

Finally, in relation to the third proposal, the FTT considered that it would be premature to determine the grouping issues in the context of the closure notice applications, because the relevant information was still incomplete “and all parties would need to prepare carefully for a full argument of the grouping test, especially as there is the distinct possibility of a referral to the ECJ being necessary or desirable”: see paragraph [72]. The FTT said again (ibid) that they considered it preferable for the issue to be raised in the context of a paragraph 31A referral to the FTT, adding that if that were not possible “then this third proposal could be raised again in any renewal of the closure notice applications”.

Following the decision of the FTT, it is common ground that the three applicants then supplied the Revenue with all the outstanding information in respect of their group structures. Nevertheless, the Revenue continued to withhold their consent to a joint referral under paragraph 31A. At first they said that they wished to wait for the forthcoming decision of the Court of Appeal in M & S (CA) II, on the ground that the need to obtain answers on the group structure questions might be affected if the decision of the Upper Tribunal were overturned in certain respects. As it turned out, however, the Court of Appeal upheld the decision of the Upper Tribunal on every point. Since then, the Revenue have continued to refuse to agree to a referral, saying that they will not agree until full replies are supplied to the Revenue’s comprehensive requests for information in relation to the no possibilities test. For their part, the applicants continue to maintain their position that it would be premature and oppressive to require them to provide all this information at the present stage of the proceedings.

In a further attempt to move matters forward, the applicants lodged renewed closure notice applications on 17 November 2011. However, those applications are again resisted by the Revenue on the basis that it remains reasonable to keep the enquiries open while the Revenue’s requests for information on the no possibilities test are still unanswered. The Revenue also argued in a letter to the FTT dated 5 December 2011, asking for the renewed applications to be dismissed, that it would in any event be premature for the FTT to make a reference to the CJEU on the group structure issues in the absence of any evidence that the no possibilities test was satisfied. Unless and until it becomes clear that the no possibilities test is met, say the Revenue, there is nothing to displace the ruling in M & S v Halsey that the group relief provisions were in all other respects compliant with EU law.

It is in the light of this procedural impasse before the FTT that the claimants enrolled in the Loss Relief GLO now ask the High Court to take the initiative by itself ordering a reference to the CJEU on the group structure issues. Meanwhile, in a further twist to the procedural warfare before the FTT, the applicants wrote to the FTT on 19 March 2012 asking for the renewed closure notice applications to be stayed pending determination of the present application by the High Court. Predictably, the Revenue have objected to the proposed stay on the basis that the application now before me is misconceived and represents a transparent attempt to circumvent the proper procedures of the FTT.

At this point it is necessary to refer to a further procedural complication. In Autologic Plc v Inland Revenue Commissioners [2005] UKHL 54, [2006] 1 AC 118 (“Autologic”), the House of Lords decided by a bare majority that the claims for cross-border group relief in the Loss Relief GLO which had been made within the usual statutory time limits for claiming group relief fell within the exclusive jurisdiction of the Special Commissioners (now the FTT), and that it would accordingly be an abuse of process if the claims were to be permitted to proceed as claims for damages or restitution in the High Court. Where, however, the claims had been made outside the prescribed statutory time limits, the Special Commissioners would have no jurisdiction to entertain them (unless the Revenue or the Special Commissioners had power to extend the normal time limits, and agreed to do so). In relation to the claims in this second category, therefore, the majority held that, subject to the claimants first requesting agreement to an extension of time for making the claims, they should proceed in the High Court.

The leading speech for the majority was delivered by Lord Nicholls of Birkenhead. In relation to claims of the second type, he said this:

“41.

In such cases the taxpayer’s remedy necessarily lies elsewhere. In such cases the taxpayer’s remedy is of a different character. The taxpayer’s remedy lies in pursuing proceedings claiming restitutionary and other relief in respect of the United Kingdom’s failure to give proper effect to Community law. The appeal commissioners have no jurisdiction to hear such claims. Such claims are outside the commissioners’ statutory jurisdiction, and the commissioners have no inherent jurisdiction. Claims in this class should therefore proceed in the High Court …

42.

I add one caveat. The revenue and the appeal commissioners have power to extend time limits for late amendments and late appeals. Before proceeding with their High Court claims claimant companies in this class of cases should therefore take the simple step of inviting the revenue or the appeal commissioners to extend the time limits appropriately. If this invitation is accepted, the claimants should proceed along the statutory route. If the invitation is declined, or if the revenue and the appeal commissioners have no power to grant the necessary extensions, the way will be clear for the High Court proceedings to continue.

43.

I recognise there may be instances where a claimant company has claims in both the classes I have described. In respect of some accounting periods a company may have made a group relief claim or still be in a position to make such a claim, in respect of more distant accounting periods it may now be too late for the company to put forward such a claim. The need for one company to pursue proceedings before the appeal commissioners and separately and additionally in the High Court is unfortunate. But this possibility is inherent in the distinction between the two classes of case: the distinction between obtaining the tax relief to which the claimant is entitled and obtaining damages for unlawful failure to make such relief available. Unless the circumstances are exceptional, having claims in both classes is not a sufficient reason for a company declining to make a group relief claim in respect of accounting periods where this can still be done.”

Lord Steyn and Lord Millett both agreed with Lord Nicholls, and Lord Millett added some observations of his own in paragraphs [62] and [63]. Powerful dissenting judgments were given by Lord Hope of Craighead and Lord Walker of Gestingthorpe.

Despite the conclusion reached by the majority, the House did not order the claims in the first category which were proceeding in the High Court to be struck out. Instead, it ordered that they should be stayed. As Lord Nicholls explained in paragraph [44]:

“The cases falling within the first class described above (“claimant companies which can still obtain group relief”) should be stayed. They should be stayed until further order rather than struck out the more readily to accommodate any unforeseen turn of events. And the stay should not preclude the court referring questions to the European Court if practical convenience so dictates. The cases in the second class (“claimant companies which cannot now obtain group relief”) should proceed in the High Court.”

To similar effect, Lord Millett said at paragraph [63]:

“It is impossible to foresee all eventualities, and I agree with Lord Nicholls that the proceedings in the High Court in respect of claims which should have been brought before the commissioners should be stayed and not struck out. This would have two advantages. It should encourage the revenue to co-operate in waiving or extending time limits and removing procedural and other obstacles to the commissioners’ jurisdiction; and it would enable the High Court claims to be revived in the event of unforeseen difficulties arising before the commissioners which cannot be overcome.”

It is clear from these passages in the speeches of Lord Nicholls and Lord Millett that the first category of High Court claims in the Loss Relief GLO remain in existence, and could in principle be re-activated by a lifting of the stay in at least two sets of circumstances. The first is an “unforeseen turn of events” before the FTT, or “unforeseen difficulties” arising before the FTT which cannot be overcome in that tribunal. The second, importantly for present purposes, is that the High Court may still refer questions to the CJEU in the context of the stayed claims “if practical convenience so dictates”.

The present position under the Loss Relief GLO is briefly as follows. At its peak, there were around 85 groups of claimant companies enrolled in the GLO. In the light of the ECJ’s judgment in M & S v Halsey, many of the claims were discontinued, including that of Autologic Plc itself (when the House of Lords heard Autologic, the Advocate General’s opinion had been delivered in M & S v Halsey but the ECJ had not yet given its judgment). As at 20 November 2012, 39 claims remained enrolled in the GLO. The great majority of the claimant groups are clients of Dorsey & Whitney (Europe) LLP, who are the lead solicitors in the GLO, although a few are clients of Pinsent Masons and RPC. In relation to the 39 claims which are still current, the Revenue contend that the group structures of the claimants fail to engage EU rights in 33 of the cases. Despite the reduction in size of the GLO from its initial peak before M & S v Halsey, it remains very substantial with hundreds of millions of pounds of tax at stake. Mr Aaronson QC emphasised, and I accept, that the claims are not going to disappear by a process of gradual attrition, and they will continue to be vigorously prosecuted by the claimants unless and until it becomes clear that, for one reason or another, they are legally or factually unsustainable.

In their skeleton argument for the hearing, counsel for the Revenue (Mr David Ewart QC, leading Ms Maya Lester and Mr David Yates) said it was the Revenue’s understanding that all of the claimants with “out of time” claims in the High Court also had “in time” claims which are within the jurisdiction of the FTT and are currently under enquiry by the Revenue. It appears, however, from a witness statement filed on 20 November 2012 by Dr Simon Whitehead of Dorsey & Whitney, that this assertion is not correct. Of the 33 groups whose group structure is in issue, there are at least six where the claimants have only “out of time” claims. Further, those six include some of the claimants in two of the most important disputed structures, namely structure (a) in paragraph 17 above and the variant of it where the common parent company is resident in another member state.

It is also apparent from the schedule exhibited to Dr Whitehead’s statement that there are several “out of time” claims where the Revenue have refused to extend the relevant time limits, or where a response to such a request is still awaited, although there are other cases where they have agreed to admit the late claims. On any view, therefore, there is a substantial number of “out of time” cases which can only proceed in the High Court, and which are not subject to any stay.

Against this background, counsel for the applicants submit that the most efficient and sensible way to resolve the stalemate before the FTT is for the High Court now to take the initiative, and to make a reference to the CJEU on the group structure questions. There is no jurisdictional impediment to such a course, because there are still several “out of time” cases which can only proceed in the High Court, and in relation to the stayed “in time” claims the present circumstances are said to provide a perfect opportunity for the kind of “practical convenience” that Lord Nicholls presciently envisaged in Autologic at paragraph [44]. The applicants accuse the Revenue of dragging their heels in the FTT, and improperly using their comprehensive requests for information on the no possibilities test as a reason for refusing to co-operate in making a reference to the CJEU on the logically prior group structure questions. Counsel submit that, in contrast with the complex factual issues which may arise under the no possibilities test, resolution of the group structure questions requires only a limited amount of information, most of which has already been supplied to the Revenue. The issues are said to be essentially ones of law, which can be debated and decided by reference to a few diagrams of relevant group structures, with virtually nothing more needed by way of evidence.

The applicants accept that the group structure issues might eventually become academic, if the Revenue were to succeed on the no possibilities test. However, it is likely to be several years before all the issues concerning the meaning and application of the no possibilities test have been finally determined, and the claimants will suffer severe prejudice if a reference to the CJEU of the group structure questions is delayed until that distant date (assuming, of course, that the claimants, or at least some of them, are ultimately successful on the no possibilities test). Counsel also point out, in this connection, that if the claimants are ultimately successful in their claims for cross-border group relief, then the longer they have to wait for their money, the more they are likely to be disadvantaged, bearing in mind that, in accordance with the decision of the House of Lords in Sempra Metals Limited v IRC [2007] UKHL 34, [2008] 1 AC 561, the claimants would be entitled to restitution measured by an award of compound interest at the rates applicable to borrowing by the Government, which are always lower than the market rates available to the claimants.

In summary, the applicants submit that the group structure issues are short and distinct points of principle that are well suited to preliminary determination; that if they are decided in the claimants’ favour, and if the claimants also succeed on the no possibilities test, there is likely to be a further delay of two to three years before they can enjoy the fruits of their success if a reference of the group structure questions is postponed; while if the claimants fail on the group structure issues, the sooner that is known the better. On this last point, although the case management of the claims proceeding in the FTT is not directly before me, and must ultimately be a matter for the FTT, the clear implication of the applicants’ submissions is that it would be unreasonable to require them to answer the Revenue’s wide-ranging requests for information on the no possibilities test before the CJEU has ruled on the structural issues. All the time and costs involved in providing answers to the Revenue’s enquiries would turn out to have been wasted if the structure of a claimant group is such that no EU rights were relevantly engaged in the first place.

I now turn to the submissions for the Revenue. In their skeleton argument, counsel put first the submission that the application amounts to an abuse of process in so far as it relates to the “in time” claims. These claims fall within the exclusive jurisdiction of the FTT, and the Revenue’s ongoing enquiries in relation to the no possibilities test relate to a matter which the claimants will have to demonstrate (the factual burden being on them) if they are to succeed in their claims. The FTT has already held, in the Finnforest case, that it is reasonable for the Revenue to continue with their enquiries at this juncture, and that it would be premature to direct the issue of closure notices. More generally, the Revenue say they are fully entitled to carry out their enquiries so that the conclusions stated in the closure notices, when they are eventually issued, will be based on as much information as it is reasonably possible to obtain: see the guidance given by the Supreme Court in TowerMCashback LLP 1 v Revenue and Customs Commissioners [2011] UKSC 19, [2011] 2 AC 457, at paragraphs [18] and [83] to [85] per Lord Hope JSC and Lord Walker JSC respectively.

In relation to the “out of time” claims, the Revenue initially contended that all the taxpayers who had such claims also had “in time” claims. As I have explained, this contention was rebutted by Dr Whitehead’s evidence produced on the eve of the hearing. That apart, however, counsel submitted that the FTT was in any event the more appropriate forum for resolution of the outstanding issues, relying on the observation of Lord Nicholls in Autologic at paragraph [21] that “detailed questions of this character are more suited for determination by the Special Commissioners than the High Court, especially where large numbers of companies are involved”. Accordingly, they submit that it makes more sense, as a matter of case management, for any “out of time” claims to await the resolution of “in time” claims by the FTT once the Revenue have closed the relevant enquiries and any appeals have been determined.

The only reason that the claimants do not find the FTT procedures to their liking, say the Revenue, is that they do not wish to go to the trouble of ascertaining whether there is any factual basis to their claims before it is decided whether M & S v Halsey applies to their group structures. Had the claimants been willing to comply with the Revenue’s requests for information in a timely fashion, they could already have gathered all the material needed to resolve the factual issues on the no possibilities test. The High Court should not now indulge the claimants for their failure to co-operate with the Revenue’s enquiries by permitting them to circumvent the procedure for determination of their claims laid down by Parliament. That procedure includes the statutory mechanism in paragraph 31A of Schedule 18 to the Finance Act 1998 for the joint referral of questions arising during the course of an enquiry to the FTT. The Revenue’s agreement is needed for any such joint referral, and the present application is again an attempt to circumvent that statutory requirement.

As an example of the Revenue’s willingness to make a joint referral pursuant to paragraph 31A of Schedule 18 in an appropriate case, counsel referred to the history of the proceedings in Philips Electronic UK Limited v HMRC (“Philips”). The basic issue in Philips was the compatibility with EU law of certain provisions of UK law relating to consortium relief within section 402(3) of ICTA 1988. The relevant claims for relief concerned the years 2001-2004, in relation to the losses of the UK branch of a company established in the Netherlands. I was informed that the claimant provided the Revenue with all necessary evidence in relation to the issues in dispute, including reports of a Dutch liquidator and expert evidence of Dutch tax law. In those circumstances, the Revenue agreed to a joint referral which came before the FTT in 2009: see [2009] UKFTT 226 (TC), [2009] SFTD 629. The decision of the FTT was appealed to the Upper Tribunal, which made a reference to the CJEU in December 2010. Advocate General Kokott gave her opinion on 16 February 2012, and the CJEU delivered its judgment on 6 September 2012: see Case C-18/11 Revenue and Customs Commissioners v Philips Electronics UK Limited, [2013] STC 41.

More generally, the Revenue submit that it is in principle unsatisfactory to refer questions of law to the CJEU before the underlying facts have been fully established. The dangers of such an approach are said to be vividly illustrated by the history of the proceedings following M & S v Halsey itself, and by the history of the litigation in the FII GLO where a further reference to the ECJ was needed in order to resolve the “corporate tree” questions which had not been dealt with on the first reference. The relatively swift and self-contained history of the Philips litigation shows, by contrast, the advantages of having a first instance decision where all disputed issues of fact and law are fully ventilated, and the court or tribunal is in a position to make an informed and complete decision as to what questions should be referred. Given the huge amounts of tax potentially recoverable by the remaining claimants in the Loss Relief GLO, it cannot be suggested that such an approach is disproportionate, or that there is anything unreasonable in the Revenue requiring full replies to their requests for information on the no possibilities test before they agree to a joint referral under paragraph 31A on the group structure issues.

Counsel also referred me in this connection to the recently published new recommendations by the CJEU to national courts and tribunals in relation to the initiation of preliminary ruling proceedings (2012/C 338/01). I will set out the passages in the recommendations which appear to me most relevant:

“The Court’s jurisdiction in preliminary rulings

…

5.

Since the preliminary ruling procedure is based on co-operation between the Court of Justice and the courts and tribunals of the Member States, it may be helpful, in order to ensure that that procedure is fully effective, to provide those courts and tribunals with the following recommendations.

6.

While in no way binding, these recommendations are intended to supplement Title III of the Rules of Procedure of the Court of Justice (Articles 93 to 118) and to provide guidance to the courts and tribunals of the Member States as to whether it is appropriate to make a reference for a preliminary ruling, as well as practical information concerning the form and effect of such a reference.

…

The decision to make a reference for a preliminary ruling

…

14.

In order to enable the Court of Justice properly to identify the subject-matter of the main proceedings and the questions that arise, it is helpful if, in respect of each question referred, the national court or tribunal explains why the interpretation sought is necessary to enable it to give judgment.

…

The appropriate stage at which to make a reference for a preliminary ruling

18.

A national court or tribunal may submit a request for a preliminary ruling to the Court as soon as it finds that a ruling on the interpretation or validity of European Union law is necessary to enable it to give judgment. It is that court or tribunal which is in fact in the best position to decide at what stage of the proceedings such a request should be made.

19.

It is, however, desirable that a decision to make a reference for a preliminary ruling should be taken when the national proceedings have reached a stage at which the referring court or tribunal is able to define the legal and factual context of the case, so that the Court of Justice has available to it all the information necessary to check, where appropriate, that European Union law applies to the main proceedings. In the interests of the proper administration of justice, it may also be desirable for the reference to be made only after both sides have been heard.

The form and content of the request for a preliminary ruling

…

24.

If it considers itself able to do so, the referring court or tribunal may, finally, briefly state its view on the answer to be given to the questions referred for a preliminary ruling. That information may be useful to the Court, particularly where it is called upon to give a preliminary ruling in an expedited or urgent procedure.”

Relying on this guidance, the Revenue submit that it is not necessary for the High Court to make a reference on the group structure issues at this stage in order to enable it to give judgment on the “out of time” claims, let alone on the “in time” claims which are proceeding before the FTT, and which would continue to proceed before the FTT after the CJEU had given its judgment. Furthermore, the need for a reference on those issues may disappear if the Revenue win on the no possibilities test.

Finally, Mr Ewart QC emphasised in the course of his oral submissions that the Revenue’s position is not that resolution of the group structure issues should be deferred until the litigation on the no possibilities test has run its full course, possibly four or five years hence. The Revenue’s position is, rather, that the claimants should comply with the normal procedure in cases which are proceeding towards a hearing in the FTT. The claimants may not want to provide the outstanding information in relation to the no possibilities test at this stage, but that is what they must do if the Revenue are to be put in a position either to issue closure notices or (if other unrelated matters remain outstanding) to make a joint referral under paragraph 31A. Either way, the matter would then proceed to a hearing before the FTT, and a fully informed decision could then be taken about what issues should be referred to the CJEU.

I was initially attracted by Mr Aaronson’s argument that the group structure issues are self-contained questions of law which the CJEU should be asked to determine while the litigation on the no possibilities test progresses towards its conclusion, and that there would be a risk of severe prejudice to the claimants if a reference on those issues were delayed until after the no possibilities test had been finally determined in the claimants’ favour. On reflection, however, I am satisfied that each limb of this argument is flawed, or at least that the position is considerably more complex than Mr Aaronson was willing to concede.

As to the first limb, I accept that the group structure issues are logically prior to the issues about the meaning and timing of the no possibilities test, and that their resolution will require a reference to the CJEU unless they are in some way rendered academic (most obviously by a conclusive decision in the Revenue’s favour on the no possibilities question). I also accept that little in the way of evidence is likely to be needed to enable the CJEU to rule on at least some of the disputed structures. But it is far from obvious to me that this will be true in respect of all the disputed structures, or that all the relevant questions could safely be decided on the basis of a short statement of agreed facts.

In particular, I agree with Mr Ewart that issues of justification and proportionality may well arise in relation to structures where the parent company is resident in another member state which are significantly different from the corresponding issues where the parent is UK resident. In my judgment there would be obvious dangers in asking the CJEU to decide those issues before full disclosure of all potentially relevant material has taken place, and perhaps before the full facts of the cases in which the reference is made have been found at trial. Furthermore, it is in my view desirable that the precise formulation of the questions to be referred should be considered, and if necessary refined, in the light of such disclosure or findings. Experience shows that questions which look deceptively simple when posed in the abstract may become far more complex and difficult, and new angles and implications may emerge, once they are put in a detailed factual context. It is a truism to say that no question of law can be decided in a factual vacuum; and even a decision which is based on a short statement of agreed facts can often turn out to be a deceptive short cut. These risks are reflected in paragraph 19 of the CJEU’s recent recommendations to national courts, with its salutary emphasis on the need to ensure that the CJEU has available to it “all the information necessary to check … that European Union law applies to the main proceedings”.

A further relevant consideration, to my mind, is the need where possible to avoid burdening the CJEU with references on questions which may prove academic. There is a clear need for national courts to exercise restraint in making references, given the greatly increased size of the Union (there are now 27 member states), the continuing growth of EU legislation, and the heavy workload of the Court with the well-known delays which it entails. The claimants do not dispute that, if the Revenue’s interpretation of the no possibilities test is correct, and is ultimately upheld either by the Supreme Court or (following a further reference) by the CJEU, then their claims could not succeed, even if the group structure issues were to be decided in their favour. This is in my opinion a powerful reason for waiting at least until the Supreme Court has ruled on the no possibilities test, or formulated the terms of a further reference to the CJEU, before making a final decision whether or not to refer the group structure questions.

Nor is this all. Under Article 267 TFEU a court or tribunal may submit a request for a preliminary ruling to the CJEU only if it considers it necessary to do so in order to resolve the dispute brought before it. Thus the High Court may make a reference where it considers it necessary to do so in order to decide a case which is pending before it, and the FTT may likewise make a reference in respect of a case before the tribunal. But what I am being asked to do, in relation to the “in time” cases, is to make a reference in the stayed High Court proceedings, not so that the High Court can then decide them, but so that the FTT can do so. It seems to me highly doubtful whether such a procedure complies with Article 267, and I think there is a real possibility that the CJEU would decline to entertain the reference.

This objection would not, of course, apply to a reference made in one or more of the “out of time” claims which are proceeding in the High Court; but to confine the reference in that way would give rise to a number of further problems. First, it is not clear to me that all of the disputed structures can be exemplified within the “out of time” claims alone. Secondly, I would be reluctant to adopt this expedient as a means of circumventing the procedural stalemate in the FTT, when the whole thrust of the majority decision in Autologic is that the appropriate forum for resolution of the claims in the Loss Relief GLO is normally the statutory appeal procedure laid down by Parliament. Thirdly, Mr Ewart told me that the “out of time” claims are likely to involve further issues of EU law, particularly in relation to time limits and the question whether the claims are implicitly excluded by the statutory scheme of group relief, which do not arise in relation to the “in time” claims. Mr Ewart submitted, and I agree, that it would be undesirable to make a reference in the “out of time” claims before those further questions had been fully investigated in their factual context, not least because this process may bring to light further questions which need to be referred to the CJEU.

I now come to the second limb of Mr Aaronson’s argument. The real problem here, as it seems to me, is the unwillingness of the claimants to provide detailed answers to the Revenue’s enquiries relating to the no possibilities test, which are of course predicated upon the law as stated by the Court of Appeal in M & S (CA) I and M & S (CA) II. To some extent, I can sympathise with this reluctance. The enquiries involved are no doubt comprehensive and often difficult to deal with, involving events which took place many years ago and in different jurisdictions. But it is the claimants who have chosen to bring their claims, involving very large sums of money, and the evidential burden lies on them to demonstrate that the no possibilities test is satisfied. The Revenue cannot reasonably be blamed for making searching enquiries when so much is at stake, and in basing those enquiries on the law as stated by the Court of Appeal, even though they no doubt hope that the test will be reformulated by the Supreme Court and/or the CJEU in a way which will ultimately make the enquiries redundant. The process may well be inconvenient, time-consuming and expensive for the claimants; but (subject to what I say below about the way forward) it is in my view a burden which they have brought upon themselves, and about which they cannot legitimately complain.

As to the spectre raised by Mr Aaronson of a further two or three years’ delay in resolution of the group structure questions if they are not immediately referred to the CJEU, I remain unpersuaded that it has any real substance. The reason for the procedural stalemate before the FTT is the unwillingness of the claimants to answer the Revenue’s enquiries on the no possibilities test. If the necessary information is provided, I see no reason to doubt Mr Ewart’s assurances that the Revenue would then co-operate in bringing appropriate cases before the FTT, either by way of appeals from closure notices or on a joint referral under paragraph 31A, with a view to the group structure issues then being referred to the CJEU. I am certainly not prepared to assume, or to find on the material before me, that the Revenue have been unreasonably dragging their heels or seeking to gain a tactical advantage from delay.

I will, however, say this. I am not sure whether the Revenue’s enquiries in relation to the no possibilities test are currently being pursued against all the claimants still enrolled in the Loss Relief GLO or only against selected test claimants. It would, I think, be unreasonable for the Revenue to insist on full disclosure from every claimant at this stage, particularly when the no possibilities test is shortly to be reviewed by the Supreme Court. If, on the other hand, the enquiries are confined to representative test cases, I do not consider it unreasonable for the Revenue to insist that they be answered. Meanwhile, however, the parties should in my view also identify a full range of suitable “in time” test cases for resolution of the group structure issues, and co-operate in bringing them before the FTT for consideration of when and in what terms a reference should be made, and of the extent to which potentially relevant facts should first be found at a hearing. In principle, I can see no good reason why these matters should not proceed simultaneously, and as far as possible it would seem to be sensible to use the same test cases both for the provision of comprehensive information on the no possibilities test and to exemplify the disputed group structures.

I appreciate that the case management of the “in time” claims is ultimately a matter for the FTT, but I also think it is appropriate for me, as the managing judge of the Loss Relief GLO, to express my views on the best way forward, albeit in the rather general terms which I have just outlined. It may well be that there are further complications or refinements which can be discussed when this judgment is handed down. The application now before me is in my judgment both premature and (at best) procedurally questionable, so it must be dismissed. But with goodwill and co-operation on both sides, it should be possible both for the claimants to provide full information about the no possibilities test in representative test cases, and for the same or other test cases to be brought before the FTT by the most convenient route in order to resolve the group structure issues.