England and Wales High Court (Chancery Division) Decisions

IN THE HIGH COURT OF JUSTICE
BUSINESS AND PROPERTY COURTS OF ENGLAND AND WALES
BUSINESS LIST (ChD)

		Royal Courts of Justice Fetter Lane, London, EC4A 1NL
		19/07/2019

Mrs Justice Falk:

Introduction

1. The Claimants, to whom I shall refer collectively as Credit Suisse, are four members of the Credit Suisse group, a financial services group headquartered in Zurich, Switzerland. The claim arises out of their payment of a tax known as bank payroll tax ("BPT") which was enacted by Schedule 1 to the Finance Act 2010 ("FA 2010") in the aftermath of the financial crisis.
2. BPT applied to "relevant remuneration" awarded by "taxable companies" during the period between the date of its announcement on 9 December 2009 and 5 April 2010 (the "chargeable period"). In very broad terms, the definition of "taxable company" covered banks, building societies and certain related entities resident or operating in the UK, and "relevant remuneration" covered discretionary payments exceeding £25,000, rather than regular salary or pre-contracted amounts. The tax was imposed on the banks rather than the employees, at a rate of 50%. In crude terms, it was a tax on bankers' bonuses. For convenience, I will refer throughout to the entities caught as "banks", although that is a significant over-simplification of the rules. As is common with tax measures, the legislation that brought BPT into effect was passed some months after the start date of the tax, when the Finance Act 2010 (the first of three Finance Acts in that year) obtained Royal Assent on 8 April 2010. Rather less usually, in this case the date of enactment also post-dated the period of operation of the tax.
3. Credit Suisse paid £238,799,581 in BPT on the due date, 31 August 2010. The final amount of BPT is not yet determined, pending the outcome of a dispute in respect of certain aspects of the rules which is currently before the First-tier Tribunal ("FTT"). By their claim to this Court, Credit Suisse allege that the imposition of BPT was unlawful on the basis that it was contrary to the State aid rules set out in Articles 107 and 108 of the Treaty on the Functioning of the European Union ("TFEU"), and they seek damages under the Francovich principle^[1] to compensate for the loss they say they have suffered.
4. The nub of Credit Suisse's case is simply stated. They say that the way in which the tax was imposed, and in particular the limited period of around four months to which it related, meant that in reality it was imposed only on those banks that as a matter of practice, policy or expectation awarded (or were practically required to award) bonuses during that period ("Taxed Banks"). Essentially, it targeted those with calendar year ends (i.e. financial years ending 31 December), bonuses generally being awarded around or relatively shortly after the end of the financial year. Banks with different year ends paid their bonuses at other times and escaped, or largely escaped, liability ("Untaxed Banks"). Credit Suisse say that this was a State aid measure because it conferred a selective advantage on Untaxed Banks, who were in a comparable legal and factual situation, and that because it was not notified to and cleared in advance by the European Commission it was unlawful.
5. HMRC say that there was no State aid or, even if there was, no "sufficiently serious" breach to found a Francovich claim, and in any event a claim for damages in respect of the tax paid is precluded by paragraph 51 of Schedule 18 to the Finance Act 1998 ("FA 1998"), as applied by paragraph 31 of Schedule 1 to the FA 2010, which requires any claim to be made pursuant to that Schedule and Schedule 1A to the Taxes Management Act 1970. Any appeal against a refusal of such a claim would lie to the FTT only, and not to the High Court. No such claim was made and the four year time limit under paragraph 31(2) of Schedule 1 has now expired.

State Aid

6. Article 107(1) of TFEU provides:

"Save as otherwise provided in the Treaties, any aid granted by a Member State or through State resources in any form whatsoever which distorts or threatens to distort competition by favouring certain undertakings or the production of certain goods shall, in so far as it affects trade between Member States, be incompatible with the internal market."

7. Article 108 relevantly provides as follows:

"1. The Commission shall, in cooperation with Member States, keep under constant review all systems of aid existing in those States…

2. If, after giving notice to the parties concerned to submit their comments, the Commission finds that aid granted by a State or through State resources is not compatible with the internal market having regard to Article 107, or that such aid is being misused, it shall decide that the State concerned shall abolish or alter such aid within a period of time to be determined by the Commission.

If the State concerned does not comply with this decision within the prescribed time, the Commission or any other interested State may, in derogation from the provisions of Articles 258 and 259, refer the matter to the Court of Justice of the European Union direct.

….

3. The Commission shall be informed, in sufficient time to enable it to submit its comments, of any plans to grant or alter aid. If it considers that any such plan is not compatible with the internal market having regard to Article 107, it shall without delay initiate the procedure provided for in paragraph 2. The Member State concerned shall not put its proposed measures into effect until this procedure has resulted in a final decision.

…"

8. In order to fall within the scope of Article 107(1), it is accepted in the light of the case law that (1) there must be aid, in the sense of an economic advantage, granted by the State or through State resources, (2) the measure must favour certain undertakings over others in a comparable legal and factual situation (a "selective advantage" must be conferred), (3) that advantage must not be justified by the nature or general scheme (or structure) of the system, and (4) it must be liable to distort competition and to affect trade between Member States. The burden of proof is on Credit Suisse^[2].

The issues

9. The parties have agreed that there should be a split trial of liability and quantum issues. This trial relates to liability only. The agreed issues for determination at this stage are as follows:

Issue 1: Did HMRC's imposition of BPT on the Claimants constitute a State aid measure within the meaning of Article 107(1) TFEU? In particular:

(a) Did the measure confer an economic advantage to the Untaxed Banks?

(b) Was the economic advantage granted by the State or through State resources?

(c) Was the measure selective?

(d) Was the measure liable to distort competition and to affect trade between EU Member States?

Issue 2: If the measure is considered to be a State aid measure, does that give rise to a claim for repayment of tax, as opposed to a recovery of aid from the beneficiaries of that aid?

Issue 4: If, by imposing BPT on the Claimants, HMRC did breach the State aid rules in Articles 107 and 108 TFEU, was that a sufficiently serious breach of EU law which renders HMRC liable to pay damages to the Claimants in accordance with the decision of the European Court of Justice in Joined Cases C-6/90 and C-9/90 Francovich and Bonifaci v Italy [1991] ECR I-5357 and subsequent case law?

Issue 8: Is the Claimants' claim (or any part thereof) precluded by:

(a) paragraph 51(6) of Schedule 18 to the FA 1998 read with paragraph 31(1) of Schedule 1 to the FA 2010? and/or

(b) paragraph 31(2) of Schedule 1 to the FA 2010?

(Issues 6 and 7 relate respectively to causation (and therefore the quantum of damages), and to the question whether Credit Suisse are entitled to interest on a compound or simple basis, and if so in what amount and for what period. Issues 3 and 5 are common ground, being respectively that the imposition of BPT was not notified in advance to and cleared by the European Commission under Article 108(3) TFEU, and that Article 108(3) has direct effect and confers rights on Credit Suisse. It is also common ground that any economic advantage was granted by the State or through State resources, Issue 1(b) above.)

The evidence

10. There were two witnesses for HMRC, Jacqueline McGeehan and Michael Williams, who provided witness statements and were cross-examined. At the relevant time both held senior roles in the personal tax team at HM Treasury, Ms McGeehan as a Deputy Director and Mr Williams as the Director of Personal Tax and Welfare Reform. Mr Williams had lead responsibility for BPT at Director level and Miss McGeehan was closely involved in its development, with responsibility for policy in relation to BPT. The evidence of both was straightforward and I accept it.
11. As discussed further below, Credit Suisse adduced no witness evidence. Documentary evidence was limited and comprised in large part materials disclosed by HMRC that were produced in connection with the development of BPT, its announcement and subsequent decisions about its scope and duration. A certain amount of correspondence between HMRC and other banks was also disclosed, together with the BPT returns of certain other banks and details produced by HMRC of BPT paid by different banks.

The background and policy rationale of BPT

12. BPT was announced by the then Chancellor of the Exchequer, the Rt. Hon Alistair Darling MP, in the Pre-Budget Report on 9 December 2009. He stated as follows:

"Supporting growth is vital to provide the future revenue to halve borrowing over the next four years, but, as I have said, it also requires us to take some tough decisions on tax now. I am determined that any tax increases will continue to be guided by our values of fairness and responsibility. Last year, the banks made collective losses of £80 billion in this country alone. This would have been much higher without the unprecedented level of support from the taxpayer. There is no bank that has not benefited, either directly or indirectly, from this help. This should be a time for banks to rebuild their capital base and become stronger. A tax on profits, as has been suggested, would prevent them from doing that, so I have decided against a windfall tax. However, there are some banks who still believe their priority is to pay substantial bonuses to some already high-paid staff. Their priority should be to rebuild their financial strength and increase their lending, so I am giving them a choice: they can use their profits to build up their capital base, but if they insist on paying substantial rewards, I am determined to claw money back for the taxpayer.

I have decided to introduce from today a special one-off levy of 50 per cent on any individual discretionary bonus above £25,000. This will be paid by the bank, not the bank employee, and anti-avoidance measures will be introduced with immediate effect. High-paid bank staff will, of course, also have to pay, as usual, income tax at their top rate on any bonus they receive. On a cautious assumption, which includes our expectation that some banks will rein back on bonuses, this levy is expected to yield just over £500 million. That additional money will be used to pay for the extra measures that I have already announced, such as help for the young and older unemployed to get back into work."^[3]

13. The Pre-Budget Report document issued on the same day stated:

"The Government attaches great importance to tackling the remuneration practices that contributed to excessive risk taking by the banking industry. The Government has made clear that the sector needs to develop sustainable long-term remuneration policies that take better account of risk and facilitate the build-up of loss-absorbing capital. However, evidence suggests that some may be intending to pay bonuses for the current year that are not consistent with a prudent approach to risk.

The Government today announces that where bank (and building society) employees are awarded discretionary bonuses, in whatever form, above £25,000 in the period from the Pre-Budget Report to 5 April 2010, the banks paying these bonuses will pay an additional bank payroll tax of 50 per cent on the excess bonus over £25,000. The tax will not be deductible in computing the taxable profits of affected companies. This tax will encourage banks to consider their capital position and to make appropriate risk adjustments when settling the level of bonus payments above the threshold which is at the level of median earnings in the UK. If banks choose to make awards that are not consistent with a prudent approach to risk, it is only fair that they contribute more to the public finances in a year when profits have been facilitated by significant taxpayer support for the banking sector as a whole.

It is intended that in the longer term, remuneration practices will be changed as a result of corporate governance and regulatory reforms, as outlined later in this chapter. The one-off bank payroll tax will apply until 5 April 2010, but the Government will consider extending the period of the charge so that the tax remains in place until the relevant provisions of the Financial Services Bill come into force. Where there is evidence of avoidance schemes being put in place, the Government will take action to close those schemes."^[4]

14. The Technical Note published by HMRC with the draft legislation, also on 9 December, states the policy rationale as follows:

"… to encourage change in the remuneration practices that contributed to excessive risk taking by the banking industry. The Government wants to encourage the development of sustainable long-term remuneration policies that take greater account of risk and facilitate the build up of loss-absorbing capital.

The bank payroll tax will encourage banks to consider their capital position and make appropriate risk-adjustments when settling the level of bonus payment this year. It is intended that in the longer term the remuneration practices will be changed as a result of corporate governance and regulatory reforms. The one-off bank payroll tax will apply until 5 April 2010, but the Government will consider extending the period of charge so that the tax remains in place until the relevant provisions of the Financial Services Bill come into force."

15. It is clear from these extracts that the overall policy rationale was one of encouraging banks to alter their remuneration practices, by taking a more prudent approach that moved away from practices that in the Government's view contributed to excessive risk taking. BPT was designed as a short-term measure to discourage banks from paying high bonuses at a time when the Government felt that they should be building up capital. Revenue raising was a secondary objective. In due course the Financial Services Bill^[5] would address the behavioural changes required via the introduction of regulatory rules governing remuneration policies, and making provision enabling certain payments to be clawed back. So BPT was intended as a short-term measure with the primary objective of encouraging bonus restraint before longer term regulatory reforms took effect.
16. The witness evidence is consistent with this, and provides further detail about the rationale for the start and end dates for the tax. As regards the start date of 9 December 2009 (the date of the announcement), any earlier date was ruled out by concerns about retrospective legislation and fairness considerations, and by the fact that the objective was to encourage a change in behaviour. It was for the same reasons that amounts that were already contracted to be paid before 9 December were excluded. Any later start date was regarded as unacceptable because of the risk that forewarned banks would bring forward bonus payments to avoid the impact of the rules.
17. The position in relation to the end date, 5 April 2010, is more complex. 5 April is of course the end of the tax year for individuals and is also an important date in determining employers' payroll tax liabilities in respect of income tax and both employees' and employers' national insurance. Income tax and national insurance rates and thresholds usually apply on a tax year by tax year basis. The understanding of those designing the tax was that the main discretionary bonus awards season was between January and March, based on banks' results for the year to 31 December. The chosen date allowed what was thought to be the main awards season to be targeted whilst ensuring that the measure was time-limited. Making the measure time-limited addressed concerns about the potential impact of BPT on the UK's international competitiveness, and recognised that the long-term solution was in the Financial Services Bill. Another highly relevant factor was the introduction, with effect from 6 April 2010, of a 50% rate of income tax for incomes over £150,000 (as compared to the then highest rate of 40%). It was thought that this would mean that employees would put strong pressure on employers not to delay bonuses. Finally, both Mr Williams and Ms McGeehan referred to impending changes to regulatory capital requirements which would potentially affect bonus levels, because of the need for banks to retain capital in anticipation of the need to meet the more stringent new rules^[6].
18. So in summary, it was recognised that going back before 9 December 2009, for example to 6 April 2009, was not an available or appropriate option, and as regards the end date the aim was to catch the main bonus season while limiting any impact on international competitiveness and recognising the expected impact of the 50% income tax rate and the new regulatory rules.
19. The evidence indicates that, prior to the announcement of the tax, there was some exploration of whether the period of the charge should be extended beyond 5 April 2010 to coincide with the regulatory reforms coming into effect, and as already indicated the announcement of BPT included a statement that the Government would consider extending the period on that basis. The point was considered further in an internal HM Treasury advice paper given to the Chancellor on 10 March 2010, which notes that an extension could be achieved by a simple amendment to the draft rules.
20. That paper explains that the original reference to the possibility of extension was intended to put off banks who might otherwise have been tempted to delay their bonus arrangements until just after 5 April, but comments that HMT was not aware of any such avoidance. It adds that the introduction of a 50% income tax rate from 6 April 2010 also meant that it was unlikely that employees would be willing to delay their bonuses. Furthermore, there was no obvious date to which the charge could be extended, given that it was not possible to predict the final form of the Financial Services Bill in view of the forthcoming general election, and extending the tax would add to fears that it was not a one-off tax, potentially damaging international competitiveness. The recommendation made was that the period should not be extended. This recommendation was agreed by the Chancellor.
21. Although the initial estimate was that the revenue raised would be of the order of £500m, it appears that the gross amount raised from BPT (ignoring behavioural effects that reduced other tax paid, and therefore affected the net amount raised) was about £3.4 billion.

Credit Suisse's submissions

22. In summary, Mr Robertson for Credit Suisse submitted that the impact of the limited chargeable period was unjustifiably discriminatory. BPT was clearly introduced in a rush, with legislative options first being considered only around a month before announcement. Although State aid was considered (and advice was taken from Mr Peretz about it), the focus of the advice obtained at the time was on the definition of the entities to which the tax applied, and there was no consideration of the impact of the limited chargeable period as between entities within the definition. The point was missed. HMRC's disclosure showed that between 5 April and 31 August 2010 HMRC had become aware that four other banks in particular, NM Rothschild, Toronto-Dominion, Royal Bank of Canada and Nomura, had avoided any substantial liability due to their non-calendar year ends. In addition, Credit Suisse believed that at least four further named banks were in a similar position. These other banks did not need to fund BPT payments, and were not constrained in their ability to pay bonuses and therefore compete for employees.
23. The equivalent French tax had not suffered from the same problem, because it had been imposed for a whole year. It was not open to banks like Credit Suisse to get around the problem by delaying payment of bonuses until after 5 April, because of the effect of the anti-avoidance rules in paragraphs 12 and 14 of Schedule 1 FA 2010.
24. Credit Suisse say that the Untaxed Banks were in a comparable legal and factual situation and had obtained an economic advantage not justified by the nature and general structure of the tax. The entire tax was unlawful. In accordance with the principle established in Laboratoires Boiron SA v Union de Recouvrement des Cotisations de Securite Sociale et d'Allocations Familiales (URSSAF) de Lyon (C-526/04) [2006] 3 CMLR 50 ("Boiron"), this was not a case where the remedy was to impose BPT on the Untaxed Banks. Credit Suisse were entitled to damages both for the direct loss of having to pay the tax and consequential losses in relation to the economic advantage conferred on Untaxed Banks and/or the prejudice to Credit Suisse's ability to compete with those banks. Their claim for consequential loss (as particularised in the pleadings in their Reply) includes a claim that BPT led Credit Suisse to reduce bonuses and increase fixed salaries and other awards at a cost that exceeded the bonus reduction, and a claim that senior employees left as a result of the bonus reduction, as well as claims for other costs, including costs arising out of the payment of BPT.

HMRC's submissions

25. Mr Peretz for HMRC submitted that there was no selective advantage, because that required the identification of the "normal" tax system and an unacceptable derogation from that, and in any event any different treatment flowed from the nature and general structure of the system. Here the "normal" system was logically BPT, but the non-taxation of bonuses outside the period of its operation (the chargeable period) could not be a derogation from BPT, since it was not the "normal" system outside that time. What Credit Suisse were seeking to do was to use a hypothetical indefinite version of BPT as the reference system, which was not permitted.
26. It was always open to a bank to make a genuine decision not to award bonuses in the period covered by BPT. The limited period of applicability reflected the objectives of the tax, such that banks not awarding bonuses during the relevant period were not in a legally or factually comparable position to those that did. Credit Suisse had also not adduced evidence to prove that an advantage had been conferred, or in relation to any impact on trade or competition.
27. Even if there was State aid, the general rule was that the Court should not award a remedy for reimbursement of the tax. Boiron provided a limited exception to the general rule which did not apply. There was also no sufficiently serious breach to engage the Francovich principle, and in any event, recovery could only be made by appeal to the FTT and was time-barred.

Issue 1: was there State aid?

Selectivity: the case law

28. The most significant dispute between the parties in relation to Issue 1 was on the question of selectivity. The leading cases relied on by both Mr Robertson and Mr Peretz were Commission v World Duty Free Group (C-20 P & C-21/15 P) [2017] 2 CMLR 22 ("World Duty Free") and Commission and Spain v Government of Gibraltar and United Kingdom (C-106/09 p & C-107/09 P); [2012] 1 CMLR 44 ("Gibraltar"), both decisions of the Grand Chamber of the Court of Justice of the EU ("CJEU").
29. World Duty Free concerned a Spanish corporate tax provision which permitted taxable undertakings to obtain a deduction for the amortisation of goodwill arising from the acquisition of a foreign company, but not a domestic company. The CJEU summarised the principles governing selectivity as follows at [54] to [60] (case references omitted):

"54. So far as concerns the condition relating to the selectivity of the advantage, which is a constituent factor in the concept of "State aid", within the meaning of art.107(1) TFEU, it is clear … that the assessment of that condition requires a determination whether, under a particular legal regime, a national measure is such as to favour "certain undertakings or the production of certain goods" over other undertakings which, in the light of the objective pursued by that regime, are in a comparable factual and legal situation and who accordingly suffer different treatment that can, in essence, be classified as discriminatory…

55. Further, where the measure at issue is conceived as an aid scheme and not as individual aid, it is for the Commission to establish that that measure, although it confers an advantage of general application, confers the benefit of that advantage exclusively on certain undertakings or certain sectors of activity…

56. As regards, in particular, national measures that confer a tax advantage, it must be recalled that a measure of that nature which, although not involving the transfer of State resources, places the recipients in a more favourable position than other taxpayers is capable of procuring a selective advantage for the recipients and, consequently, of constituting State aid, within the meaning of art.107(1) TFEU. On the other hand, a tax advantage resulting from a general measure applicable without distinction to all economic operators does not constitute such aid…

57. In that context, in order to classify a national tax measure as "selective", the Commission must begin by identifying the ordinary or "normal" tax system applicable in the Member State concerned, and thereafter demonstrate that the tax measure at issue is a derogation from that ordinary system, insofar as it differentiates between operators who, in the light of the objective pursued by that ordinary tax system, are in a comparable factual and legal situation…

58. The concept of "State aid" does not, however, cover measures that differentiate between undertakings which, in the light of the objective pursued by the legal regime concerned, are in a comparable factual and legal situation, and are, therefore, a priori selective, where the Member State concerned is able to demonstrate that that differentiation is justified since it flows from the nature or general structure of the system of which the measures form part…

59. Further, it must be recalled that the fact that only taxpayers satisfying the conditions for the application of a measure can benefit from the measure cannot, in itself, make it into a selective measure…

60. It follows from all the foregoing that the appropriate criterion for establishing the selectivity of the measure at issue consists in determining whether that measure introduces, between operators that are, in the light of the objective pursued by the general tax system concerned, in a comparable factual and legal situation, a distinction that is not justified by the nature and general structure of that system…"

30. As can be seen from this, there are a number of elements to selectivity. In particular:

a) it must be determined whether the measure favours certain undertakings (or sectors) over others;

b) those undertakings must be in a "comparable factual and legal situation";

c) whilst a tax advantage can constitute State aid, it will not do so if it results from a general measure applicable without distinction to all economic operators;

d) the starting point is to identify the ordinary or "normal" tax system, and then determine whether the tax measure is a derogation from that system;

e) the question whether undertakings are in a comparable factual and legal situation must be determined in the light of the objective pursued by the ordinary tax system; and

f) even if a measure is a priori selective, there will be no State aid where the Member State shows that the differentiation in treatment flows from the "nature or general structure" of the system.

31. The CJEU went on at paragraphs [71] to [79] to consider its earlier judgment in Gibraltar. That case related to a new corporate tax system introduced in Gibraltar, comprising a payroll tax, business property occupation tax ("BPOT") and a registration fee. Liability to payroll tax and BPOT were capped at 15% of profits. The requirement to make a profit, and capping tax at 15% of profit, were not found to confer selective advantages, because the measure applied without distinction to all economic operators (paragraph [83] of the judgment). However, the imposition of a payroll tax and BPOT was found to be selective, since both taxes inherently favoured offshore companies with no real physical presence in Gibraltar.
32. In World Duty Free the CJEU explained that the measure considered in Gibraltar did not take the form of a tax advantage that derogated from an ordinary tax system, but rather involved the application of a general tax scheme based on general criteria. However, selectivity extended to measures which by their effects favoured certain undertakings, in that case offshore companies, on account of their specific features. There was de facto discrimination between entities that were in a comparable position in the light of the objective of the regime, which it found to be the objective of "putting in place generalised taxation of all resident companies" (paragraph [74]). Selectivity could be established if the measure was an "integral part" of an ordinary tax system rather than a derogation from it, and it was "sufficient, in order to establish the selectivity of a measure that derogates from an ordinary tax system, to demonstrate that that measure benefits certain operators and not others" where those operators are in a comparable situation in the light of the objective of the system (paragraph [76]). Whilst it was not always necessary that a tax measure should derogate from an ordinary tax system, the fact that it can be so characterised is "highly relevant" where the effect of the measure is to subject two categories of operator to different treatment, namely those falling within the derogating measure and those falling within the ordinary tax system. However, the key question was whether the effect of the measure, irrespective of its form or the means used, placed recipient undertakings in a more favourable position (paragraphs [77] to [79]).
33. Read with the decision in Gibraltar, what the CJEU was saying was that the important point is to focus on the effect of the rules, rather than their form or, for example, the drafting technique chosen. Whilst State aid in a tax context is most typically granted in the form of an exemption or relief from a tax charge that would otherwise apply (that is, a derogation), it can still arise if the measure is not structured in that way but has that effect. So, for example, in Gibraltar at [87] the Court stated that it had consistently held that what is now Article 107(1) distinguishes between measures by reference to their effects, independently of the techniques used (referred to as "regulatory techniques" in the following paragraphs of the judgment). In that case the General Court had fallen into error by focusing on the technique used and concluding that the system in question had not been designed to include derogating provisions. The CJEU commented on that as follows:

"92. Such an interpretation of the selectivity criterion would require, contrary to the case law cited in [87] above, that in order for a tax system to be classifiable as "selective" it must be designed in accordance with a certain regulatory technique; the consequence of this would be that national tax rules fall from the outset outside the scope of control of State aid merely because they were adopted under a different regulatory technique although they produce the same effects in law and/or in fact.

93. Those considerations apply particularly with regard to a tax system which, as in the present case, instead of laying down general rules applying to all undertakings from which a derogation is made for certain undertakings, achieves the same result by adjusting and combining the tax rules in such a way that their very application results in a different tax burden for different undertakings."

34. However, in Gibraltar the Court also emphasised the importance of determining a reference framework. Whether there is an advantage can only be established by a comparison with "normal" taxation (paragraph [90]). The existence of a selective advantage entailed mitigation of the charges "normally" included in the budget of an undertaking (paragraphs [71] and [89]). The reference framework in that case was identified as the tax regime at issue, which applied to all undertakings (paragraph [95]). That regime in practice discriminated between companies in a comparable situation with regard to the objective of the rules, which was to introduce a general system of taxation for all companies established in Gibraltar (paragraph [101]). The non-taxation of offshore companies was the inevitable consequence of the way in which the bases of assessment were designed (paragraph [106]).
35. It can be seen from this that the "normal" or "reference" framework was determined having regard to what was perceived to be the objective of the regime, in that case to introduce a general corporate tax regime. Against that reference point, the non-taxation of offshore companies was in effect a derogation from the "ordinary" system, being one of subjecting companies to taxation.
36. Mr Peretz also relied on a recent case of the General Court, Poland v Commission (T-836/16 and T-624/17) ECLI:EU:T:2019:338 ("Poland"), released on 16 May 2019. No official English translation of the decision is available, so what follows relies on a translation commissioned by HMRC.
37. Poland related to a turnover tax imposed on retailers, which the Commission challenged under State aid rules on the grounds that the progressive rates provided for (involving both zero and lower rates) impermissibly benefited companies with low turnover, as compared to those required to pay tax at the highest rate. The General Court annulled the relevant decisions. It referred to the need to identify the "normal" tax regime, which in that case would relate to the retail sector only, and held at paragraph [65] that the "level of the levy, like the tax base, the taxable event and the field of taxable persons, forms part of the fundamental characteristics of the levy's legal framework". What the Commission had done was to identify a hypothetical "normal" system where all retailers' turnover was taxed at a single rate, whereas what was required was to determine selectivity by reference to the actual characteristics of the normal tax regime, taking account of its progressive structure (paragraphs [66] to [68]). Furthermore, the Commission was wrong to conclude that the progressive structure ran contrary to the objective of the tax and thus had discriminatory effects. The Court found that the objective was to introduce a sectoral tax in accordance with the principle of tax redistribution. Its progressive nature was consistent with that objective, namely to introduce redistributive sectoral taxation on retailers' turnover. There was no selectivity where differences in treatment did not run contrary to the objective and were simply the result of the application of the "normal" regime, provided that the relevant advantage was not arbitrary, was applied in a non-discriminatory manner and did not negate the objective of the tax (paragraphs [70] onwards, in particular paragraphs [73], [83], [89], [91] and [94]).
38. Mr Peretz also referred to GIL Insurance Ltd v Commissioners of Customs and Excise (C-308/01) [2004] 2 CMLR 483 ("GIL"), a case relating to the introduction of differential rates of insurance premium tax (IPT). In that case the CJEU considered whether the application of a higher rate was justified by the "nature and general scheme" (structure) of the system. The higher rate had been introduced to counteract value shifting between the price of retail goods and associated insurance, and the Court found that it was not intended to confer an advantage on operators who continued to be subject to the standard rate. Standard rate IPT was not a derogation from the general system of taxation of insurance, or a tax scheme favouring a specified sector. So even on the assumption that there was a selective advantage, it was justified by the nature and general scheme of the system. (See paragraphs [73] to [78].)
39. It is worth noting that the objective of the relevant tax regime is relevant both to the exception for measures justified by the nature and structure of the scheme and to identifying whether entities are in a comparable situation. Its relevance to the former is clear from the approach adopted in GIL. Mr Peretz suggested that on the facts of this case there was little difference between these two tests.
40. Following the hearing, the General Court released another decision with similarities to Poland, Hungary v Commission (T-20/17, released 27 June 2019) ("Hungary"). An official English language version is available. This case also concerned sectoral taxation on turnover, with progressive tax rates. As in Poland the Court found that the Commission had impermissibly hypothesised a system with a single rate, and that it had erred in not recognising that the progressive rate structure was consistent with the redistributive objective of the tax. At paragraph [77] the Court also commented that the "nature" of the system must be identified by reference to its objective, in the sense of the "basic or guiding principles", and at [101] that there is no selectivity if differences stem from the straightforward application of the "normal" system and the variation mechanisms do not misconstrue the objective of the tax. The Court's findings are in my view entirely consistent with the cases referred to by Counsel, and so I did not request any further submissions relating to Hungary.

Selectivity: discussion

The reference system

41. As World Duty Free makes clear, the starting point in determining whether a tax measure is selective is to identify the "normal" tax system. This is the reference point. Mr Robertson says that the "normal" system was the corporate (presumably including payroll) tax system as it applied to banks at the relevant time, in the absence of BPT.
42. This seems to me to give rise to a problem. What the case law requires is the identification of the normal system and then a determination of whether the tax measure is a derogation from that, either by design or (as illustrated by Gibraltar) in its effect. On my reading of the cases, derogation is used in the sense of benefit or advantage. For example, in World Duty Free the Court explained at paragraph [74] that the Court in Gibraltar "… held that the nature of that scheme did not preclude a finding that the measure concerned was selective … since the condition relating to selectivity has a broader scope that extends to measures which, by their effects, favour certain undertakings …", and at paragraph [76] "… it is sufficient, in order to establish the selectivity of a measure that derogates from an ordinary tax system, to demonstrate that that measure benefits certain operators and not others …".
43. Interpreting derogation as meaning a benefit or advantage is of course also consistent with the fact that what the rules capture are impermissible selective advantages granted through State resources.
44. The difficulty with using "no BPT" as the reference system is that the Untaxed Banks had no advantage as compared to the Taxed Banks, because their treatment was no better with BPT in place than it was under the reference system.
45. Mr Robertson's response to this is that all that is required is a differentiation between the two classes of banks. But that seems to me to pay insufficient attention to the case law, which explains that in a tax context a reference point of a normal system is required, in order to determine whether a selective advantage is conferred. Gibraltar does not support Credit Suisse's case because, properly understood, it does not have the effect that it is not necessary to show that any advantage is conferred as compared to the normal system, but only that all that must be shown is that the measure has that effect.
46. Mr Robertson suggested that Boiron illustrated the point he was making. However, Boiron (discussed below) is an authority on the remedies available, not the question of whether State aid existed. That issue was dealt with in the earlier related decision of Ferring SA v Agence Centrale des Organismes de Securite Sociale (ACOSS) (C-53/00) [2003] 1 CMLR 34. It is interesting that, although in Boiron the Court described the measure at issue as not involving an exemption from tax, in Ferring that is precisely the way in which the non-application of the charge at issue to some undertakings was described: see paragraphs [18] and [20]. I therefore did not find these cases to provide any support to Mr Robertson's submissions on this issue.
47. An alternative reference system is of course one with BPT in place. The reference system might be BPT itself, or a wider system of taxation of banks (whether covering payroll related taxes, or extending more broadly, for example to corporation tax). However, BPT applied equally to Taxed Banks and Untaxed Banks for the period to which it applied: there was no differentiation. Before 9 December 2009 and after 5 April 2010 BPT did not apply, and therefore again there was in fact no differentiation. Any assumption that BPT applied outside its actual period of operation would fall foul of the error made by the Commission in the Poland case, namely to hypothesise a system that did not exist.
48. I agree with Mr Peretz that, once these points are taken into account, what Credit Suisse are essentially left with is a challenge to the start and/or end dates of the regime.

Challenge to start and end dates

49. The thrust of Credit Suisse's case is that they would have had no complaint if BPT had applied for 12 months, because then all banks would be treated in the same way and could not escape the tax due to the date to which they happened to draw up their accounts. What they are complaining about is the effect of its brief period of operation. They never quite pinned their colours to the mast of any particular 12 month period to which they said the tax should have applied. I got the impression that Credit Suisse's main focus was on a failure to extend the tax after 5 April 2010 (presumably to 9 December 2010 or some subsequent date), rather than having an earlier start date. This may reflect the fact that the primary objective of BPT was to dissuade banks from awarding large discretionary bonuses, an objective which would not be met by seeking to apply it to bonuses awarded prior to announcement, as well as the objections (whether legal or political in nature) to retrospective legislation. But it may also reflect the challenges involved in reaching a conclusion that banks awarding bonuses before 9 December 2009 were (retrospectively) to be regarded as receiving State aid, once BPT was announced and implemented, despite their inability to identify that fact at the time of the awards.
50. However, the evidence indicates that HM Treasury's focus was in fact on the tax year to 5 April 2010 and on the proportion of bonuses awarded in that year that might be caught by the measure, having regard to the retrospectivity issues that had to be addressed. The period after 5 April 2010 was of concern in relation to steps banks might take to defer bonuses into that period, rather than in relation to later bonus awards more generally.
51. Mr Peretz submitted that, in principle, changes in tax law and decisions about the dates on which such changes should have effect are matters for Member States. Requiring BPT to remain in force would conflict with the legislature's right to change domestic tax rules (including, in this case, following the May 2010 general election), and its short-lived nature was consistent with its objectives. Mr Robertson's response was that discretion over tax policy was constrained by the State aid rules. A new Government could have revoked the whole tax rather than keep it in force, so that banks were not required to pay BPT on 31 August 2010. The publicly stated objective was to rein in bank bonuses, and this did not just apply to banks with a 31 December year end. The objective of the tax had to be judged by its declared purpose. The tax was in place for an unprecedented short period of time.
52. It is obviously correct that the State aid rules limit Member States' discretion over domestic rules. I also doubt that a situation could never arise in which the start or end dates of a tax resulted in State aid arising. Examples might arise where, for example, it is obvious from the dates chosen and the factual circumstances that particular operators are being singled out for favourable treatment as compared to comparable operators. But, as discussed further below, whether there is selectivity must be judged in the light of the objectives of the relevant measure. Those objectives can be determined in the normal way, with evidence, and contrary to Mr Robertson's submission that I should, I do not see that I must limit myself to (for example) published objectives incorporated in the legislation, without regard to the witness and other evidence available.

The objectives of the reference system, and whether Taxed Banks and Untaxed Banks were in a comparable situation

53. There was no evidence or submissions about the objectives of any part of the tax system apart from BPT, so I will proceed on the basis that insofar as the relevant reference system is one without BPT, the objective of that system is the standard one of generating revenue for the Exchequer.
54. As regards BPT, as already explained it was designed primarily to discourage banks from paying high bonuses, with revenue raising as a secondary objective. I have found that BPT was intended as a short-term measure, with other regulatory changes expected to address the required behavioural changes in the longer term. Those designing the tax intended it to capture what they understood to be the main bonus awards season in the early part of 2010.
55. It is clear from the evidence that it was understood that not all bonuses would be covered, whether because they had already been paid before 9 December 2009, or had been contractually agreed before then, or were (genuinely) awarded after the period for which BPT applied. However, there is no evidence that HM Treasury or HMRC gave conscious thought to the position of banks who did not, or did not generally, pay bonuses during that period because they did not have calendar year ends. It was understood that banks generally had 31 December year ends and awarded bonuses around that time. It was also understood that because bonuses awarded before announcement would not be covered, a proportion of bonuses awarded in the tax year 2009-10 that would otherwise fall within BPT would not be caught.
56. By reference to the objectives of discouraging high bonuses and revenue raising, Taxed Banks and Untaxed Banks were in a comparable situation. The Government was concerned about the culture of high bonuses throughout the industry. While it was in operation, BPT applied on its terms to all banks.
57. The limited period that the tax was in operation was such that BPT would have had a different impact on a bank with a practice of paying bonuses in that period, as compared to a bank with no such practice, at least to the extent that there were commercial or other constraints on it altering that practice in a way that meant that it fell outside the charge to BPT. That limited period of operation was a deliberate feature of the regime: it was intended as a short-term measure, capturing the main bonus season for 2009-10, the view being taken that thereafter other developments, including regulatory developments and the changes to income tax rates, would mean that there was no longer the same need for BPT.
58. Whilst it was a deliberate feature, it is not immediately obvious that it would be correct to describe the limited period of operation as part of the objectives of BPT. It might be said that the start and end dates simply reflected decisions about the earliest date from which the objectives could start to be implemented, and about when those objectives were regarded as sufficiently met, taking account of other policy considerations.
59. However, I am persuaded that that would be too narrow an approach. BPT was, as one witness put it, a "stop gap". It was fundamental that it applied to a short period, before longer term changes were introduced: its objective was to tax bonuses in the period to which it applied. It was a one-off tax. The focus was on bonus awards in the tax year 2009-10, but with recognition that there were sound reasons why bonuses awarded before announcement could not be caught. Extension of the charge after 5 April 2010 was initially kept open as a possibility, but that was to deter possible avoidance activity in 2009-10.
60. Looked at that way, and taking account of the reasoning behind the start and end dates, Taxed Banks and Untaxed Banks were not in a comparable situation. The main objective of the tax was to deter bonuses being awarded during the 2009-10 tax year. That objective could by definition not be met in respect of bonuses awarded before the date of announcement: such bonuses, and the banks awarding them, were in a different position from banks that had yet to award bonuses. As regards bonuses (genuinely) awarded after 5 April 2010, they were never the target of the measure. After that time there was a new income tax rate in force, and regulatory changes were looming. I do not think it matters that the regulatory changes were not in force until the end of 2010. They were viewed as sufficiently proximate to mean that, taken together with other factors, BPT was no longer required. Therefore banks awarding bonuses after 5 April 2010 were, again, not in a comparable position with those awarding them during the chargeable period for BPT.

Was it a general measure applicable without distinction?

61. For the period of its operation, BPT applied without distinction to Taxed Banks and Untaxed Banks. The complaint is that the effect of the limited period was to create a distinction in application.

Difference arising from the "nature or general structure"?

62. This point only strictly arises if there is selectivity in principle. In the same way as in GIL, where the rules in question were not intended to confer an advantage on operators who continued to be subject to the standard rate, the BPT regime was not intended to confer an advantage on banks whose bonuses were not caught by BPT because they fell outside its period of operation. Rather, it was intended to tax banks that fell within its terms.
63. In my view the "nature and structure" of BPT includes the fact that it was intended to apply, and did apply, for a short period (consistent with its objectives). That was inherent in the regime. On that basis any differentiation in treatment between Taxed Banks and Untaxed Banks flows from that nature or general structure and is consistent with it, such that there could be no State Aid.

Conclusions on selectivity

64. In order to establish selectivity, a reference ("normal") system must be established and an advantage demonstrated against that normal system. Credit Suisse's case that the reference system is one with no BPT creates a difficulty in demonstrating any advantage to Untaxed Banks. The real challenge must be to the start and end dates of the regime.
65. The objectives of BPT included that it was to apply for a short period, related to the tax year ended 5 April 2010. In the light of its objectives, Taxed Banks and Untaxed Banks were not in a comparable situation. The imposition of BPT on Credit Suisse was therefore not selective for State aid purposes.

Economic advantage/impact on trade and competition: evidence

66. A notable feature of this case was that Credit Suisse adduced no witness evidence or substantive documentary evidence to demonstrate, as they were required to do to establish State aid, that an economic advantage was in fact conferred on at least one Untaxed Bank, and that that advantage was liable to distort competition and affect trade between Member States. For example, they provided no witness evidence to support their case that, because the group had a calendar year end, it was in practice required to pay bonuses during the period of operation of BPT, whereas comparable banks with different year ends were not, and thereby obtained an advantage. There was also no evidence about specific bonuses paid by other banks at times before or after the period of operation of BPT.
67. The evidence that was available included the accounts of some banks containing limited financial information about bonus awards, dates to which various banks' accounts were drawn up and the amounts of BPT paid by a list of banks. Credit Suisse relied primarily on HMRC's own evidence, including the evidence of HMRC's witnesses, to demonstrate that there was a bonus paying season in the early part of the year, related to the fact that most banks have calendar year accounting periods, and submitted that it could be inferred from the substantial amount of BPT paid by those banks that they did not have a practical choice to avoid paying bonuses during that period.
68. Credit Suisse's case is that they and other banks in their position were compelled in practice to pay bonuses while BPT was in force, because they always did that and they had to meet employees' expectations, and they say that that is supported by evidence provided by HMRC. That evidence included some limited references to advice that had been provided by Slaughter and May to HM Treasury about bonus practices during the development of the tax. However, those designing the tax in HM Treasury and HMRC were clearly working on the basis of a pretty limited, and incomplete, understanding of the position, and certainly without reference to the specific positions of Credit Suisse or any Untaxed Banks.
69. Credit Suisse no doubt had reasons for choosing not to adduce witness or direct documentary evidence on this topic. It is inconceivable that they would have been unable to do so as far as their own position was concerned. Essentially, they are asking the Court to draw inferences on matters relating to their own affairs that would ordinarily be proved as primary facts in the normal way, through witness and documentary evidence.
70. There is clearly some additional complication in relation to the position of other banks. A reluctance on the part of other banks to provide information about their own bonus practices and the impact of BPT is understandable in any circumstances, even ignoring the State aid context, although HMRC remarked that Credit Suisse had made no effort to require information to be provided using third party disclosure orders. Again, Credit Suisse would have had their own reasons for not pursuing that, but it does not assist the Court. In the case of the Untaxed Banks, Credit Suisse are asking the Court to draw inferences from the amounts of bonuses shown in those banks' accounts and the levels of BPT paid by them without, for example, any clear confirmation that the bonuses were discretionary in nature or were of the requisite size to have been caught by BPT were it in force. The best evidence available comprises some limited references in correspondence between the four banks named at [22] above and HMRC, in which the banks indicate that they consider that they do not have any, or any material, BPT liability because of the timing of bonus payments. This is very limited evidence on which to conclude that there was at least one other bank that conducted a comparable business at the relevant time, paid material bonuses and escaped BPT as a result of its limited duration.
71. These points are not simply technical or procedural. It is worth bearing in mind that a finding by this Court that State aid exists has implications beyond the case in question. The Commission's notice on the enforcement of State aid (OJ 2009/C 85/01) requires a national court confronted with unlawfully granted aid to draw all legal consequences, including "in principle" ordering full recovery from the beneficiary (paragraph 30). In practice, of course it would be up to HMRC to take steps to recover aid so far as they were able to do so, but the point is that there would be an obligation to take action. It is not open to Credit Suisse to say that they are looking for a finding that BPT was unlawful, rather than a conclusion that State aid was conferred on any other bank. They cannot have one without the other.
72. Given my conclusion that there was no selectivity, it is neither necessary, nor in my view appropriate, to make findings of fact based on inferences, as would be required to demonstrate the existence of an economic advantage in this case.
73. A similar point arises in relation to the requirement to demonstrate that any advantage conferred was liable to distort competition and trade. Since it is not necessary to prove an actual impact on competition or trade, the "bar" is somewhat lower than it is with the need to demonstrate economic advantage, so the point has somewhat less force in this context. Nevertheless, the burden of proof is on Credit Suisse. They say that the Court can draw inferences from the fact that banks operating in the City of London conduct business in or with other Member States under the EU regime for financial services businesses (so that there was a potential impact on trade), and point to the Government's concern that an extension of BPT might have affected international competitiveness as proving that the measure was liable to affect competition. Credit Suisse also rely on concerns expressed to the Court during these proceedings by two of the banks named at [22] above in relation to disclosure and confidentiality, which refer to Credit Suisse being a "competitor bank". As regards competition, I should record that I agree with HMRC that the evidence relied on in relation to the potential impact on international competitiveness does not by itself prove that the test is met. It was clear from the witness evidence that the Government's concern related to the potential adverse effect of the continued operation of BPT on the competitiveness of the UK as a location to conduct banking business. That is not the same thing as saying that any advantage conferred on Untaxed Banks by failing to impose BPT for an entire year distorted, or threatened to distort, competition. The direct evidence available is the brief reference in the correspondence to the Court. In my view further inferences would be required to reach conclusions about the impact or potential impact on competition and trade, and I do not propose to attempt to draw them.

Conclusion on issue 1

74. In conclusion on issue 1, Credit Suisse have not established a selective advantage, and so the imposition of BPT did not constitute a State aid measure within Article 107(1) TFEU.
75. This conclusion strictly makes it unnecessary to consider any of the remaining issues, but I will consider Issues 2 and 4 because they were fully argued before me, and in case I am wrong in my conclusion on Issue 1. As discussed below, Issue 8 raises a point that has recently been considered in another case.

Issue 2: claim to repayment of tax?

76. Where State aid is established in relation to a tax measure, the general rule is that non-aided undertakings are not able to recover tax that they have paid but from which others have been relieved, on the basis that the effect of allowing them to do so would be to extend rather than correct the aid. As the CJEU stated in H. J. Banks & Co. Limited v The Coal Authority and the Secretary of State for Trade and Industry (C-390/98) [2001] 3 CMLR 51 at [80]:

"Persons liable to pay an obligatory contribution cannot rely on the argument that the exemption enjoyed by other persons constitutes State aid in order to avoid payment of that contribution."

77. Banks also makes it clear that this is the case even if the aid cannot in fact be recovered from the beneficiary.
78. However, there is an exception to this general rule, which was established in Boiron. The scope of this exception was helpfully discussed in an opinion of Lord Burns in the Outer House of the Court of Session in Cloburn Quarry Co Ltd v HMRC [2013] CSOH 203, [2014] STC 1073. Using the paragraph numbering from the opinion as issued by the Court and reflected in the Simon's Tax Cases report (there is some inconsistency between those and other reported versions) he outlined the position at [32] as follows:

"… art.107(1) of the TFEU prohibits selective aids not a selective imposition of charges. Thus exemptions to tax might be viewed as unlawful and selective state aid but the charge itself would not. However, if the charge or tax itself is used to finance state aid, then the tax or charge itself might be held to be unlawful state aid. There is therefore an exception to the general rule if there is what is termed 'deliberate asymmetric taxation' where a state chooses to tax one group of undertakings while not imposing the tax on a competing group with the objective of distorting competition in favour of the non-taxed groups. In that situation the asymmetric imposition of tax is not the result of the aid but the means by which the aid is granted."

79. He explained the position further at [44] as follows:

"It can be readily appreciated that, in terms of the general rule, a person is not entitled to withhold a tax or a levy on the basis simply that others are exempted from paying it and those exemptions are, or may be, unlawful and represent state aid. … That general rule is a powerful one and is not easily displaced. However, it is clear that there are exceptions to it. The exception said to exist here was identified in the opinion of the Advocate General [Tizzano] in Air Liquide Industries Belgium SA v Ville de Serang and Province de Liège (Joined cases C-393/04 and C-41/05) [2006] ECR I-5293, para 70 of the opinion at note 29 as follows: 'The answer would have been different had the imposition of the tax itself been unlawful, insofar as it was designed to create an unlawful situation benefiting certain persons, as in Boiron in which I today delivered my opinion… That case concerns a tax established in the form of an asymmetric charge, since it is designed solely to apply to certain economic operators and not to others in a competitive relationship with the former, and is introduced specifically and only to create a situation that favours the undertakings which are not subject to it. There is thus a close link between the tax and the aid, like two sides of the same coin, because the advantage given to the undertakings which are not subject to the tax exactly corresponds to the disadvantage imposed on the undertakings liable to it. In a situation of that kind, therefore, it is the actual imposition of the tax which may be deemed to be unlawful in the light of the Community rules on aid.'"

80. At [45] Lord Burns explains that the Advocate General defined an asymmetric charge at paragraph 38 in his opinion in Boiron as meaning a charge imposed only on some economic operators but not on others in competition with them, in order to offset costs alleged to be borne by the latter. Boiron was a case involving only two sets of operators, wholesale pharmaceutical distributors on whom the State had imposed certain public service obligations, and pharmaceutical laboratories who sold the same products and were not subject to such obligations. The charge was imposed on the latter and not the former, and it was the measure imposing the tax that was found to constitute the aid, rather than any separate exemption. At [49] and [50] Lord Burns said that it was readily ascertainable in that case that the two groups of operators were in direct competition, and that the charge was specifically designed to compensate the group on which the obligations had been placed. In contrast, in the case he was considering (relating to Aggregates Levy), the levy was not imposed "specifically and only" to create a situation favouring undertakings not subject to it, but instead primarily to reduce perceived environmental damage. His reasons for distinguishing Boiron included the lack of any primary objective of correcting a disadvantage or compensating competitors, his conclusion that the Aggregates Levy was not an asymmetric charge in the sense of one imposed on some operators to offset costs borne by others, and the fact that the tax in Boiron was not a tax of general application in which exemptions amounted to the aid measure (paragraphs [52] to [55]).
81. I agree with this assessment of Boiron. The Advocate General described that case as relating to an "entirely special situation" in which the principal aim of the charge was to compensate for an alleged imbalance between two groups of undertakings, and the charge was intended to give a competitive advantage to competitor undertakings not liable to it by means of its asymmetrical imposition (paragraphs 39 and 41 of his opinion in Boiron). He went on to draw an analogy with Belgium v Van Calster (C-261/01) [2004] 1 CMLR 18, which established that where State aid takes the form of subsidies funded by a charge introduced specifically for that purpose (a "parafiscal" charge), the method of financing being an integral part of the measure, repayment of the charge should be ordered because it was only by that means that the pre-existing situation could be restored. Similarly, on the facts of Boiron a "convincing and necessary link (if not complete identity)" could be perceived between the charge and the advantage: they were inseparable aspects of the measure, and furthermore the quantum of the aid necessarily depended on the amount of the charge (paragraphs 45 to 47).
82. The CJEU agreed, explaining at [34] and [35] that the aid derived from the fact that there was direct competition between operators liable and not liable for the charge, and that the absence of liability for the latter was a deliberate if not principal objective of the tax. The imposition of the charge would "constitute an act giving effect to an aid measure", although only to the extent that the imposition of liability led to overcompensation for the additional costs borne in discharging the public service obligations ([37]). The measure conferring aid was not an exemption that was separable from the tax, and reimbursing the tax would not allow the national court to increase the number of recipients of the aid, but rather was the appropriate way of limiting the anti-competitive effects of the measure ([39] to [41]). The charge and alleged aid were even more closely linked than in the case of the parafiscal charge considered in Van Calster ([45]).
83. It is clear from Banks and Boiron that a key underlying principle is that any remedy awarded should not have the effect of extending or increasing aid, and that a charge will only be refunded if that is the appropriate way of limiting the aid. That may be the case where the tax and aid go hand in hand: in essence in Boiron the imposition of the tax charge not only itself constituted the aid, but it was designed with that objective. Similarly in Van Calster the charge was designed to and did fund the subsidies.
84. That is not the case here. It was no part of the objective of BPT to provide an advantage to Untaxed Banks. Whilst those banks may have been better off through not having to pay BPT, that is no different to the normal situation to which the general principle referred to in Banks applies. Refunding the tax would not ameliorate any aid, but extend it.
85. Accordingly, if State aid was established then it would not give rise to a claim for repayment of tax, and I would therefore decide Issue 2 in favour of HMRC.

Issue 4: Francovich: sufficiently serious breach?

86. HMRC did not dispute that if State aid was established then, notwithstanding the answer to Issue 2, this issue would still arise in respect of any consequential losses suffered by Credit Suisse.
87. In determining whether any infringement of Article 108(3) is sufficiently serious to give rise to a claim for damages under the Francovich principle, a multi-factorial assessment must be undertaken. The criteria to take into account and weigh up were summarised by the Court of Appeal in R. (on the application of Negassi) v Secretary of State for the Home Department [2013] EWCA Civ 151; [2013] 2 CMLR 45 at [14] as:

"…(1) the importance of the principle which has been breached; (2) the clarity and precision of the rule breached; (3) the degree of excusability of an error of law; (4) the existence of any relevant judgment on the point; (5) whether the infringer was acting intentionally or involuntarily or whether there was a deliberate intention to infringe as opposed to an inadvertent breach; (6) the behaviour of the infringer after it has become evident that an infringement has occurred; (7) the persons affected by the breach or whether there has been a complete failure to take account of the specific situation of a defined economic group; (8) the position taken by one of the Community institutions in the matter".

88. In Byrne v Motor Insurers Bureau [2009] QB 66 at [45] Carnwath LJ noted that the "sufficiently serious" test is not hard-edged. It requires a value judgment by the national court, taking into account the relevant factors.
89. Clearly, the State aid rules are important. There is also no discretion conferred on Member States in relation to compliance with them. As regards criteria (2) to (4), it is worth remarking that the limited period for which a tax is in force has not previously been held to be State aid. It is a novel argument, which was presumably not immediately apparent to anyone (bearing in mind that Credit Suisse's own claim was only made in August 2016)^[7]. Although the BPT rules were announced with immediate effect, rather than after prior consultation, that was done for obvious reasons relating to forestalling risk, and there was a period following announcement when representations were considered and various changes were made to the scope of the rules to take those representations into account. This point was not raised, and the evidence indicates that it was only after the end date of 5 April had been confirmed, and the Finance Act 2010 had been enacted, that the four banks named at [22] above contacted HMRC to explain that they considered that BPT did not apply to them or had limited effect.
90. As regards criterion (5), whilst the legislation was obviously intentional, and resulted from deliberate policy choices, there was clearly no deliberate intention to infringe, and indeed advice was taken from experienced Counsel about the possible State aid risk in respect of other aspects of the rules, without the point being picked up. Importantly, there was clearly no intention to confer an advantage on Untaxed Banks, unlike, for example, the facts of R. v Secretary of State for Transport Ex p. Factortame Ltd (C-48/93) [1996] 1 CMLR 889, where the Government's intention was to protect fishing communities.
91. Mr Robertson submitted that Mr Peretz's advice (privilege in respect of which had been waived) was that the State aid position was not risk free, and therefore the proper course was to notify the Commission. I disagree. First, the point does not assist Credit Suisse because the risks being considered related to other aspects of the rules. Secondly, as confirmed by Ms McGeehan's evidence the Government takes a risk-based approach in relation to State aid, and will not notify a measure just because it is told that there is some element of risk. That seems to me to be a reasonable and proportionate approach to take where substantial risk is not identified. Ms McGeehan's recollection was that they had concluded based on the legal advice, and taking account of the rationale for the tax, that notification was not required.
92. As regards (7), there was no failure to take account of the position of the Taxed Banks: banks paying bonuses during the period of operation of BPT were the target of the measure. As regards (8), it appears that Credit Suisse have not complained to the Commission. In any event the Commission has expressed no view.
93. In all the circumstances I would conclude that, if there were State aid, there would be no "sufficiently serious" breach to justify an award of damages for consequential losses, and would therefore decide Issue 4 in favour of HMRC.
94. An additional point, although I do not need to decide it, is as follows. It strikes me as a somewhat strange result to conclude that the principle in Banks prevents Credit Suisse from recovering the BPT they have paid (as Credit Suisse accepted it would if Boiron did not apply), but that they are still entitled to recover for consequential losses, at least to the extent that those losses relate closely to the tax paid, such as funding the costs of payment, and the impact of payment on the group's regulatory capital position. I can see that the position might differ in respect of other losses, such as in respect of a loss of competitive advantage related to a reduction in bonuses.
95. The point was not addressed orally, but in written submissions Mr Robertson and Mr Singla referred in particular to Transalpine Ölleitung in Österreich GmbH v Finanzlandesdirektion für Tirol (Case C-368/04) [2007] CMLR 19 at [56] to [57]. In that case the CJEU referred to the importance of national courts taking particular care to ensure that aid is not extended (the Banks principle), whilst acknowledging that they might have to rule on applications for compensation for damage caused. In the later case of Centre d'Exportation du Livre Francais (CELF) v Societe Internationale de Diffusion et d'Edition (SIDE) (C-199/06) [2008] 2 CMLR 20 the Court recognised at [53] and [55] that a national court might be required to uphold claims for compensation for damage caused by unlawful State aid. The Commission's notice on the enforcement of State aid refers to these cases at paragraphs 43 and 44 and suggests the importance of the ability to bring such a claim as a means of direct financial compensation. However, my attention was not drawn to any authority where damages have been awarded for breach of State aid rules, and I note that the leading text book authority, Bacon (European Union Law of State Aid, 3^rd ed.) records no such decision (paragraph 20.31).

Issue 8: paragraph 31 Schedule 1 FA 2010/paragraph 51(6) Schedule 18 FA 1998

96. As with Issue 2, this issue relates to Credit Suisse's claim to recover damages in respect of the tax repaid, rather than consequential losses. The parties agreed that the two sub-issues in Issue 8, relating to the effect of the exclusion of other remedies in paragraph 51(6) Schedule 18 FA 1998 and the time limit under paragraph 31(2) Schedule 1 FA 2010, stand or fall together. If paragraph 51 prescribes the sole remedy then the four-year time limit under paragraph 31(2) will apply.
97. Given my decision on the other issues, I do not need to decide Issue 8. A very similar point was recently considered by the Chancellor of the High Court in Claimants Listed in Class 8 of the Group Register of the CFC and Dividend GLO v HMRC [2019] EWHC 338 (Ch), where he concluded at [119] that common law claims for both restitution and for damages under the Francovich principle were precluded by paragraph 51(6) Schedule 18 FA 1998, on the basis that that provision did not fall foul of the principle of effectiveness. That decision is the subject of an appeal, although I understand that the appeal is currently stayed. In the circumstances I consider it preferable not to express a view on Issue 8.

Conclusion

98. In conclusion, I decide Issues 1, 2 and 4 in favour of HMRC. The claim fails.

Note 1   Joined cases C-6/90 and C-9/90 Francovich and Bonifaci v Italy [1991] ECR I-5357.    [Back]

Note 2   Other than in relation to issue (3), where it has been held that it is for the Member State to establish this justification: Netherlands v Commission (C-159/01) at [43].    [Back]

Note 3   Hansard, House of Commons debates 9 December 2009, column 367.    [Back]

Note 4   2009 Pre-Budget Report, page 44.    [Back]

Note 5   This Bill was presented to Parliament on 19 November 2009 and eventually became the Financial Services Act 2010, section 6 of which introduced section 139A of the Financial Services and Markets Act 2000. This provision (since replaced) required the Financial Services Authority (FSA) to make rules in respect of remuneration policies. The relevant rules made came into effect from 1 January 2011.    [Back]

Note 6   The relevant Directive, 2009/111/EC, was adopted in September 2009 and became effective on 31 December 2010.    [Back]

Note 7   Mr Robertson suggested that Italy v Commission (T-211/05) [2009] ECR I-2777, referred to in Bacon (European Union Law of State Aid, 3rd ed.) at 2.124, was an example of limited duration. But that related to a measure that provided tax reliefs to companies that were newly listed during a specified 15 month period, rather than to a tax that was in force for a limited period. The reliefs were clearly available to some companies and not others, and in practice only to those that were already in the process of preparing for listing, or were ready to do so. The regime also favoured Italian registered companies. (See paragraphs [120] to [122].)    [Back]