The Illusion of Motion: Corporate (Im)Mobility and the Failed Promise of Centros

The European Court of Justice’s landmark decision in Centros was heralded as creating the preconditions for a vibrant market for incorporations in the EU. In practice, however, today’s corporate landscape in Europe differs little from that of the late 1990s. Very few large companies have made use of their ability to subject themselves to the company law of a Member State in which they are not also headquartered, and there are few signs suggesting that a ‘European Delaware’ will emerge in the near future. To the extent that Member States have engaged in competitive law-making, this has largely been confined to minimum capital requirements and rules affecting the ease of the incorporation process—areas concerning primarily micro-companies. We argue that the modest effect of Centros is not only a function of limited economic incentives to engage in regulatory competition and regulatory arbitrage, but also of the fact that the applicability of large sections of relevant laws governing corporate behaviour is determined by real seat-like connecting factors which render regulatory arbitrage more difficult. We analyse the boundaries between the lex societatis and neighbouring legal areas, notably insolvency and tort law, and find that the body of rules regulating a company’s outward-facing activities, as opposed to its internal affairs, is largely removed from regulatory arbitrage. It therefore seems likely that the potential benefits of selecting the applicable company law, while remaining subject to a cocktail of other, equally relevant rules, are sufficiently small to be regularly outweighed by the costs of a complex and non-standard corporate structure that is necessary to exercise free movement rights.


Introduction
It has now been 20 years since the European Court of Justice delivered its landmark Centros decision, 1 followed by a string of cases similarly supportive of corporate mobility. 2 In theory, this line of cases has created significant opportunities for companies and entrepreneurs throughout the EU to engage in regulatory arbitrage in relation to the company law rules they are governed by. This could well have prepared the ground for Union-wide regulatory competition between Member States. In practice, however, today's corporate landscape in Europe differs little from that of the late 1990s. Very few large companies have in fact adopted structures under which the core corporate law rules they are governed by are supplied by a Member State in which they are not headquartered, 3 and both regulatory arbitrage by entrepreneurs and competitive law-making by Member States has largely been confined to the areas of minimum capital requirements 4 and the ease of incorporating (micro-)companies.
Fears of, or hopes for, the emergence of a European Delaware have thus been shown to be almost entirely unfounded, with corporate mobility and regulatory competition in EU company law-for better or worse 5 -never having reached a point warranting comparisons to the experience in the United States. 6 From this perspective, the effect of Centros on company law and company law-making in Europe was ultimately very modest.
There are several reasons for the differences between corporate mobility and the choice of law in Europe and the US. One is rooted in economics: the economic incentives for EU Member States to engage in regulatory competition are very different from those in the US, where attracting incorporations (historically New Jersey and Delaware, more recently also Nevada) offers more direct and obvious financial advantages. 7 Without disputing the importance of these incentives as an explanatory factor, this article will argue that the structure and interplay between supranational EU law and national law, in particular an idiosyncratic combination of primary EU law, conflict of law rules, and national corporate, insolvency, tort law and other legal areas, has contributed significantly to the generally muted response to Centros by both 1 Case C-212/97 Centros Ltd v. Erhvervs-og Selskabsstyrelsen [1999] ECR I-1459. 2 Gerner-Beuerle et al. (2018a). 4 See e.g. Gelter (2005); Ringe (2013). 5 We do not offer an opinion on the merits and demerits of regulatory competition and the related debate about a race to the top or bottom in the United States. The classic exposition of the 'race to the bottom' view is Cary (1974), and the most influential response is Winter (1977). 6 For an early accurate prediction of this outcome, see e.g. Enriques (2004). For recent evidence of a lack of regulatory competition, see e.g. , Part 1, paras. 90-93. 7 On the limited importance of (hypothetical) franchise tax revenues and revenues from legal and financial services in the EU, see Gelter (2005), pp 259-262. companies and Member States. The extraordinary attention that the Court of Justice's quintuplet Centros, Überseering, Inspire Art, Cartesio and VALE has received from legal commentators has arguably distracted from the fact that the degree of corporate mobility and regulatory competition is a function of a complex interaction of legal rules of different provenance operating at different levels.
It is worth remembering that Centros and its progeny merely require Member States to accept companies incorporated in other EU Member States establishing economic ties, including their strongest economic ties, 8 in their territories, without calling into question the general applicability of the foreign company law in question. Importantly, however, this case law does not delineate the exact scope of the foreign rules to be accepted by the host Member State, or place limits on the host state's ability to enact non-company law requirements which fulfil purposes similar or have effects comparable to those of core company law rules. The ambiguity is perhaps clearest at the intersection of company law and insolvency law, but as will be shown below goes well beyond it.
In this article we argue that within the legal framework created by Centros, the potential benefits of corporate mobility as a way to select only corporate law rules, while remaining subject to a cocktail of other, equally relevant rules, are sufficiently small to be regularly outweighed by the direct and indirect costs that a complex and non-standard legal structure inevitably entails. Moreover, in the few areas where the potential benefits of exercising this choice could have warranted significant movement, defensive company law harmonisation has effectively ensured that companies and their shareholders are unable to derive meaningful benefits from exercising free movement rights. As these factors are unlikely to change significantly in the immediate future, it is unlikely that effective regulatory competition between Member States will emerge. Section 2 maps the continuing influence of real seat-like connecting factors in the legal framework governing companies, examining company law and neighbouring areas. We argue that the regime regulating a company's outward-facing activities, as opposed to its internal affairs, is largely removed from regulatory competition and arbitrage. Section 3 explores whether adherence to real seat-like connecting factors in national law is in compliance with primary EU law, in particular the right of establishment under the Treaty on the Functioning of the European Union (TFEU). Section 4 concludes.

Choice of Law
Meaningful regulatory competition in corporate law requires companies to be able to effectively choose the legal rules governing them without this choice necessitating other changes to the company's operations. In this sense, regulatory competition in corporate law relies on a fiction: the election of the applicable corporate law based on a connecting factor that can easily be manipulated and must, therefore, largely be fictitious or meaningless, as opposed to being grounded in an economic and social reality. The paradigmatically meaningless connecting factor is, of course, the registered office, at least where it merely indicates the jurisdiction of incorporation and thus determines the applicable law according to the incorporation theory, which in its purest form simply provides that a company is governed by the company law of the country where it is incorporated. In all matters falling within the remit of corporate law (for purposes of private international law), the forum accordingly applies its own law if a company is, or seeks to be, incorporated under the forum's law, and it recognises a company as a legal entity governed exclusively by foreign corporate law if it has been validly incorporated under the law of a foreign country, without imposing aspects of its internal corporate law on that company. 9 In this pure form, the incorporation theory cannot be found in any jurisdiction, 10 including the UK, which is often referred to as a quintessential incorporation theory country. All jurisdictions apply certain aspects of their internal corporate law to foreign companies, for example in order to protect third parties in their dealings with the company. In the UK, examples of such rules include transparency obligations imposed on foreign companies and the extension of the directors' disqualification regime to companies with a 'sufficient connection' with the UK. 11 These rules concern questions that are typically addressed by a jurisdiction's company law, namely 11 The Companies Act 2006 (ss. 1044-1059) and the Overseas Companies Regulations 2009 (SI 2009/1801) require companies incorporated outside the United Kingdom that open an establishment in the UK to register as an overseas company and deliver to the UK registrar a return including information on the company's name, legal form, registration in its country of incorporation, directors and secretary, the extent of the powers of the directors or secretary to represent the company in dealings with third parties, and particulars regarding the UK establishment. Furthermore, the UK rules on directors' disqualification apply to the directors of 'any company which may be wound up under […] the Insolvency Act' (Company Directors Disqualification Act 1986, s. 22(2)), which captures any company that has a 'sufficient connection' with the UK, for example because assets are located within the UK and the UK courts have jurisdiction over persons interested in the distribution of the assets (Insolvency Act 1986, s. 221(5) (c)). In other Member States, the incorporation theory is sometimes explicitly qualified through subsidiary connecting factors that incorporate certain elements of the real seat theory, even though most Member States have adopted a form of incorporation theory that applies generally to EU-incorporated companies, both companies whose registered office is located in the forum and companies that are incorporated elsewhere in the EU (which goes beyond the requirements of Centros). For example, France and Portugal allow third parties to rely on the law at the place where the company's centre of administration is located, if this is not the incorporation law; for details, see , Part 1, para. 59. 9 Technically, it would not be a deviation from a pure form of incorporation theory if a country's private international law provided that the location of the registered office determined the applicable company law, and its internal company law required the real seat of the company to be located within the country's territory. However, the nature of such a requirement as determining the applicable law (and hence qualifying as a conflict rule) or being a matter of substantive company law once the applicable law has been determined is not always self-evident (on this question, see Gerner-Beuerle and Schillig (2010), pp 317-318, discussing a provision of Hungarian law stipulating that the Hungarian Company Act should only govern companies 'which have their seat in Hungary', Art. 1(1) of Law No. CXLIV of 1997 on Commercial Companies). For purposes of regulatory competition, it is irrelevant whether a real seat requirement was qualified as a conflict rule or an element of internal company law. The possibility to compete for incorporations would be severely impeded if the domestic law provided that the location of the real seat within the forum's territory or a similar corporate connection was a precondition for incorporation under domestic law. 10 For a coding of the 'pureness' of the incorporation theory across Member States (though without considering the relationship to other areas of law) see Gerner-Beuerle et al. (2018a), p 22. the information that must be disclosed about a company that trades with third parties and the eligibility requirements of persons who serve on corporate boards.
Ultimately, the regulation of a question as a part of company law or another legal area is a function of somewhat arbitrary policy decisions. While there is widespread consensus across legal systems that the body of rules regarded as 'company law' regulates core aspects of a company's existence, notably its formation and internal governance structure, 12 other aspects, in particular concerning a company's dealings with third parties, are governed by company law in some, and by adjacent legal areas in other jurisdictions, or, more typically, they are governed by a combination of interrelated company-law and non-company-law institutions in one and the same legal system. If characterisation for purposes of private international law mirrors these conceptual distinctions, the domain of the incorporation theory may be limited, and it will certainly not represent a comprehensive set of rules regulating a company's existence and business activities. 13 Consequently, the relevance of the registered office as a connecting factor is subject to a twofold qualification. First, it is not applicable where a legal question is not characterised as falling within company law for purposes of private international law. Secondly, even where it falls within the scope of the lex societatis (from the perspective of the lex fori or from a comparative perspective, depending on how characterisation is performed), 14 the forum (or host state) which seeks to regulate the activities of a company incorporated elsewhere may elect to apply individual rules or sets of rules belonging to its company law or to another legal area that would not be applicable by virtue of general, in the EU often uniform, conflicts rules to the foreign company. We term this latter avenue to an application of host state law regulation by means of 'outreach statutes'. In the next two sections, we will first discuss two uniform conflict rules that have the effect of limiting the scope of the lex societatis, before we come to outreach statutes in Sect. 2.3.
Corporate mobility and regulatory competition are thus impeded to the extent that subject matters relevant to companies and their dealings with third parties are removed from the lex societatis or, in spite of belonging to the lex societatis, are governed cumulatively by the lex causae and certain outreach statutes of the lex fori, and the connecting factor that applies in lieu of or in addition to the registered office is 'grounded in reality', that is, it cannot be manipulated as easily as a fictitious connecting factor such as the registered office. 15 The relevant connection with the territory of a host state will vary in intensity depending on the Member State and the type of legal mechanism in question and may range from the domicile of parties to an agreement or the location of assets or business activities of some significance in the host state to a real seat equivalent (COMI). We will now turn to the question whether, outside of the lex societatis, such connecting factors 'grounded in reality' apply, and what the remaining reach of the incorporation law is where they exist.

Real Seat Equivalent: Insolvency Regulation
The closest connection of a company's actual activities with a particular territory recognised in private international law for the purpose of determining the applicable law is represented by the location of the real seat or, similar but not identical, the company's centre of main interest (COMI). What constitutes the 'real seat' of a company is not defined uniformly across jurisdictions, but usually it is the place of central decision-making, that is, the place where the board of directors regularly meets. 16 The Insolvency Regulation defines COMI as 'the place where the debtor conducts the administration of its interests on a regular basis', which must be 'ascertainable by third parties'. 17 These two connecting factors, of course, will not necessarily lead to an application of the law at the place that is most affected by the company's business activities, since the place of central decision-making may be different from the place where the company trades, has production facilities, or raises financing. However, they are a proxy that, it is presumed, will generally ensure that the law of the country in which the centre of a company's activities is situated applies. At least historically, before the advent of modern communication technology and fast travelling, the real seat will have been a fairly effective proxy of the country most closely connected to a company's activities.
The application of that state's law, it is further argued, is justified, because by deciding to pursue business activities predominantly in a particular state, the incorporators have accepted the policy choices made by the domestic legislator, as expressed in binding laws regulating, for example, the company's capacity, internal organisation, the liability of the corporate organs, and the protection of creditors, 18 who expect an application of the local law. 19 The fact that the real seat theory relies 15 A survey of legal practitioners confirms that corporate mobility is impeded in practice, see Gerner-Beuerle et al. (2016), pp 87-88 (a clear majority of 59.5% of respondents of the survey state that there is legal uncertainty as regards the boundary between the applicable company law and other areas (e.g., insolvency, tort, and contract law); notably, respondents from traditional real seat countries observe such legal uncertainty). 16 NJW 1967, 36, 38;BGH RIW 2000, 555, 556;Kindler (2018), para. 412 (with further references). 19 See the Insolvency Regulation, recital 28: '[S]pecial consideration should be given to the creditors and to their perception as to where a debtor conducts the administration of its interests.' In the 2015 Insolvency Regulation, the criterion of ascertainability by third parties is an element of the very defini-on the place of central decision-making, rather than the centre of a company's business activities, and the existence of a rebuttable presumption of identity between the registered office and the centre of main interest in the Insolvency Regulation, 20 do not warrant a different assessment of the policy rationale underpinning the real seat theory and the Insolvency Regulation, but simply serve to facilitate the identification of the 'most affected state' 21 with a view to increasing legal certainty and reducing transaction costs. 22 Whether it is convincing to justify an application of the law at the real seat with the hypothetical will of the incorporators who 'opt into' that law by locating the company's main activities there 23 is questionable, especially after Centros and the Court's acknowledgement that the Treaty grants a right to choose the law by which a company is governed. 24 Reliance on the 'perceptions of creditors' 25 as part of the definition of COMI also sits uneasily with the general approach of the Court to creditor protection in the context of corporate mobility, which is based on the premise that creditors do not require protection going beyond another Member State's company law, provided they are 'on notice' that a company has been incorporated under foreign law. 26 Nevertheless, the conflict rules of the Insolvency Regulation introduce a clear element of a 'real' connection between a company and the applicable law into the European framework on corporate mobility.
The scope of the lex concursus is set out in a non-exhaustive list in the Insolvency Regulation. It provides that the lex concursus shall determine the conditions for the opening of insolvency proceedings, their conduct and closure, and further enumerates a number of issues falling within the scope of international insolvency law. 27 These questions are mostly concerned with the operation and effects of the insolvency proceedings themselves, so that the problem of an encroachment on matters that may be regarded as 'belonging' to company law does not arise. However, the demarcation between company law and insolvency law is less clear with regard to legal mechanisms intended to protect creditors before a company is actually insolvent, in particular where it concerns acts in the vicinity of insolvency that jeopardise the creditors' interests or aggravate a company's insolvency. The most important of such mechanisms are a shift of directors' duties to creditors in the vicinity 27 Insolvency Regulation, Art. 7(2). tion of COMI. This notion codifies previous case law of the European Court of Justice, which stressed that the criterion of ascertainability was paramount, see   Kindler (2018), para. 412. 22 This is confirmed by the interpretation of the rebuttable presumption of Art. 3(1) Insolvency Regulation by the CJEU, see Eurofood (n. 19), paras. 33-35. 23 See BGH NJW 1967, 36, 38. 24 Centros (n. 1), para. 27. 25 See n. 19 above. 26 Centros (n. 1), para. 36. of insolvency; the duty to recapitalise (or liquidate while still solvent) a company; the liability of directors for a failure to protect the assets of a company when the company nears insolvency or after cash-flow insolvency or over-indebtedness, such as wrongful trading, a failure to file for the opening of insolvency proceedings, and action en responsabilité pour insufissance d'actif (liability for insufficiency of assets); the liability of directors or shareholders for causing the company's insolvency or frustrating claims of creditors; the re-characterization of shareholder loans given, or not called in, when the company nears insolvency; and avoidance actions. 28 Some of these mechanisms are structurally designed and internally classified as company law, for example a shift of directors' duties or the duty to recapitalise, and others as insolvency law, for example liability for insufficiency of assets. However, they seek to address the same economic problems: asymmetric information between creditors and the company and the misalignment of incentives, which result in a mispricing of the risk of asset substitution and the misallocation of resources, because directors and shareholders pursue non-value-maximizing investment projects in the vicinity of insolvency. 29 If these mechanisms are characterized functionally for purposes of private international law, they should accordingly all fall within the scope of either the lex societatis or the lex concursus. However, since characterization is governed by the lex fori and the dividing line between insolvency law and company law is difficult to define without reference to the boundaries drawn by the internal laws of a jurisdiction, it is not surprising that Member States have developed mutually incompatible and inconsistent solutions and legal uncertainty has until recently been high. 30 The situation has somewhat improved following a line of preliminary ruling requests that have given the Court of Justice an opportunity to decide on the characterisation of several of the above legal mechanisms and develop general criteria that can guide the delimitation of international company law and insolvency law. The rulings centre on an interpretation of what is now Article 6(1) Insolvency Regulation, which provides that the courts of the Member State where insolvency proceedings have been opened also have jurisdiction 'for any action which derives directly from the insolvency proceedings and is closely linked with them'. 31 The Court's interpretation of 'closely connected action' within the meaning of the Insolvency Regulation mirrors its approach to construing the bankruptcy exception in the EU Judgments Regulation (now the Recast Brussels Regulation), which excludes 28 For an overview of the distribution of these mechanisms in the EU, see Gerner-Beuerle and Schuster (2014), pp 302-305. The mechanisms are analysed in greater detail in the country reports in the appendix to . 29 See Davies (2006), pp 306-307;Eidenmüller (2006), pp 241-244. 30 See the text at nn. 64-72 below. 31 Art. 6(1) codifies the decision of the Court of Justice in Case C-339/07 Christopher Seagon v. Deko Marty Belgium NV [2009] ECR I-767, para. 21, which concerned an avoidance action pursuant to the German Insolvency Code, s. 129. The Court held that the Member State where insolvency proceedings were opened had international jurisdiction 'to hear and determine actions which derive directly from those proceedings and which are closely connected to them'. Except for a brief reference in recital 6, connected actions were not regulated by the predecessor of the 2015 Insolvency Regulation, Regulation (EC) No. 1346/2000 on insolvency proceedings. 'proceedings relating to the winding-up of insolvent companies […] and analogous proceedings' from the scope of the Regulation. 32 In Gourdain v. Nadler, 33 the Court of Justice held that the exception applied to actions that 'derive[d] directly from the bankruptcy or winding-up and [were] closely connected with [insolvency] proceedings'. 34 In the case at hand, which concerned the French action en comblement de passif (the liability of a director or de facto director who committed a 'management fault' that contributed to a shortfall in the company's assets), 35 the Court ascribed importance to the fact that the action was brought 'on behalf of and in the interest of the general body of creditors […] [in derogation] from the general rules of the law of liability [of directors]' and the application to the court could only be made by the liquidator. 36 These three factors-a derogation from the common rules of civil and commercial law, the protection of the interests of the general body of creditors, and an action brought by the liquidator-have been reiterated and amplified by the Court in more recent decisions arising under the Insolvency Regulation. They seem to be the guiding principles that determine the scope of the insolvency court's international jurisdiction and, as will be discussed presently, also the scope of the lex concursus, which is the law of the Member State where insolvency proceedings are opened. 37 Before we come to the applicable law, however, we will offer some comments on the three conditions that must be met for an action to be regarded as 'closely connected'.
An action derogates from the common rules of civil and commercial law if it 'finds its source' in rules that are 'specific to insolvency proceedings', and not in general civil or commercial law. 38 This formulation is ambivalent, but the Court's case law on connected actions makes it clear that the Court does not have the 'source' of a rule in a Member State's domestic law in mind, for example in the Member State's company law or insolvency legislation, but the circumstances under which a rule applies. 39 The clearest cases are provisions that require that insolvency 32 Regulation (EU) No. 1215/2012 on jurisdiction and the recognition and enforcement of judgments in civil and commercial matters [2012] OJ L351/1, Art. 1(2)(b). The Court of Justice regards the scope of the Insolvency Regulation and the Recast Brussels Regulation as mutually exclusive and exhaustive. Art. 6(1) Insolvency Regulation and the bankruptcy exception of the Recast Brussels Regulation 'must be interpreted in such a way as to avoid any overlap between the rules of law that those texts lay down and any legal vacuum', C-157/13 Nickel and Goeldner Spedition GmbH v. 'Kintra' UAB, ECLI:EU:C:2014:2145, para. 21. Therefore, interpreting either provision necessarily determines the scope of the other. 33  H v. H.K., ECLI:EU:C:2014:2410, concerning provisions of German law imposing liability on directors or managers of public and private companies for payments to creditors (or other conduct that impairs the assets of a company) at a time when the company is cash-flow insolvent or over-indebted, but before insolvency proceedings have been opened. The liability provisions are laid down in s. 64 sentence 1 Limited Liability Companies Act [Gesetz betreffend die Gesellschaften mit beschränkter Haftung] and proceedings have been opened, such as the French action en responsabilité pour insufissance d'actif, discussed in Gourdain v. Nadler, 40 wrongful or fraudulent trading, 41 or actions to set aside a transaction detrimental to the interests of the creditors entered into during a certain number of years before the opening of insolvency proceedings (avoidance actions). 42 More widely, the Court held in H v. H.K. and Kornhaas that provisions that sanction the conduct of a director at a time when the company was insolvent, without insolvency proceedings having (yet) been opened, may be closely connected, including where the action can be pursued by creditors outside of insolvency proceedings. 43 Finally, while not decided by the Court in H v. H.K. and Kornhaas, it may be argued that conduct that does not occur while a company is insolvent, but in the vicinity of insolvency, may in appropriate circumstances give rise to a closely connected action. For example, under German law, directors are liable to a company if they make payments to shareholders that cause the insolvency of the company. 44 The proximity that must exist between the payment and the company's insolvency according to the German provision is comparable to the requirements of some other legal mechanisms that impose liability on directors for acts in the vicinity of insolvency, such as wrongful trading, which is characterized as insolvency law by the UK courts. 45  and Kornhaas (n. 39), a director is liable to pay compensation to a company to the extent that the director makes payments after the company has become cash-flow insolvent or over-indebted (the liquidation value of an asset received as consideration by the company may be subtracted from the compensation claim, Müller (2018), § 64, paras. 149-149g). While the company's claim is typically enforced by the liquidator after insolvency proceedings are opened, the opening of insolvency proceedings is not a precondition for the provision to apply, Müller, ibid., paras. 173-174 (similar considerations apply to s. 92(2) sentence 1 Stock Corporation Act, Fleischer (2019), § 92, para. 37). If the court refuses to open insolvency proceedings because the assets are insufficient, creditors may be entitled to pursue the action individually (see Müller,ibid.,para. 174). In this case, the action would no longer qualify as a connected action pursuant to the third condition of Gourdain v. Nadler, see the text at nn. 51-52 below and Müller,ibid.,para. 176. 44 Limited Liability Companies Act, s. 64 sentence 3, and Stock Corporation Act, s. 92 (2)  will not be able to satisfy its debts as they fall due, unless the director, exercising the care of a prudent businessman, could not have known that this would be the consequence of the payment. 46 As with the liability provision at issue in H v. H.K. and Kornhaas, the company can claim compensation from the director, irrespective of whether or not insolvency proceedings have been opened, although the claim will typically be brought by the liquidator after the company has gone into insolvent liquidation. Given these similarities, it seems likely that the Court of Justice would qualify the claim as a closely connected action. 47 The second condition of Gourdain v. Nadler is satisfied if an action benefits all creditors, notably by increasing the funds available for distribution, for example, as a result of holding a director or shareholder personally liable. 48 Whether this condition has much independent relevance is questionable, since claims of an insolvent company will always benefit the creditors as a general body if they are enforced successfully. In order to determine whether or not a particular claim can, in principle, fall within the scope of the Insolvency Regulation, the first condition is a more useful screening mechanism, since claims that do not arise in derogation from the 'common rules of civil and commercial law', as defined above, will not be connected with the insolvency of a company. 49 If this first condition is satisfied, it only remains to be seen whether the claim is actually enforced in the context of insolvency proceedings (see the third condition, discussed in the next paragraph) in order to determine whether it falls within the scope of the Insolvency Regulation. Through this lens, the reference by the Court of Justice to 'the interest of the general body of creditors' can therefore be understood as a clarification of the other two conditions that is largely absorbed in them and illustrates how insolvency law may be differentiated from company law in functional terms. 50 46 BT-Drs. [Documentation of the Federal Parliament] 16/6140, 47. Compare this formulation with s. 214(2) UK Insolvency Act 1986 (wrongful trading). 47 This is also the opinion of some commentators, see for example Fleischer (2019), para. 41a; Krawczyk-Giehsmann (2019), pp 18-20. 48 Gourdain v. Nadler (n. 33), p 744. 49 Examples are Nickel and Goeldner (n. 32) (see the discussion in the text at n. 56 below); C-292/08 German Graphics Graphische Maschinen GmbH v. Alice van der Schee [2009] ECR I-8421 (holding that an application for the adoption of protective measures based on a retention of title clause that was brought by a seller in the insolvency of the buyer was not a closely connected action, but constituted 'an independent claim, as it [was] not based on the law of the insolvency proceedings and require[d] neither the opening of such proceedings nor the involvement of a liquidator', para. 32); C-641/16 Tünkers France v. Expert France, ECLI:EU:C:2017:847 (concerning an action for damages for unfair competition brought by the subsidiary of an insolvent company); and C-535/17 NK v. BNP Paribas Fortis NV, ECLI:EU:C:2019:96 (holding that a tortious claim for damages brought by a liquidator in the interest of all creditors against a third party that had allegedly acted wrongfully towards the creditors (a so-called Peeters/Gatzen claim under Dutch law) was 'based on the ordinary rules of civil and commercial law and not on the derogating rules specific to insolvency proceedings', para. 34). 50 It is of course not possible to distinguish insolvency law from company law solely on the basis of the function of a provision as protecting 'the interest of the general body of creditors'; otherwise much of company law would have to be qualified as insolvency law for purposes of the Insolvency Regulation and private international law. For this reason, the Court's assessment of whether a rule derogates from the common rules of civil and commercial law goes beyond an identification of the protected interests, see text at nn. 39-44 above.
Turning to the third condition, the involvement of a liquidator, the case law of the Court of Justice initially seemed to imply that an action was closely connected with insolvency proceedings if it could only be brought by the liquidator in the event of the insolvency of a company. 51 In later cases, the Court clarified that an action that is 'actually brought in the context of insolvency proceedings' is a closely connected action (provided the other two conditions are satisfied), even if it concerns a claim that does not require the opening of insolvency proceedings and can also be enforced outside of insolvency proceedings. 52 Conversely, if it is brought as an individual action and not in the context of insolvency proceedings, it falls within the scope of the Recast Brussels Regulation. 53 This is also the case if an action that can generally only be pursued by the liquidator has been assigned to a third party for a percentage of the proceeds obtained from the claim as consideration. 54 The consequence of this distinction is that many legal mechanisms have a dual nature. They are characterized as insolvency law for purposes of international jurisdiction 51 For example, in Gourdain v. Nadler (n. 33), pp 744-745, the Court emphasised that it was 'only the "syndic" In this case, a Swedish company had suspended payments and the district court at the place of the company's seat had issued a company reconstruction order ('företagsrekonstruktion'), pursuant to which the company's creditors were paid part of their claims. The balance of the claims was acquired by an investment company, which then brought a claim against one of the directors of the insolvent company for the outstanding amount. The action was based on a provision of Swedish company law pursuant to which directors are personally liable for the debts of a company if they allow the company to continue to trade at a time when it no longer has sufficient funds and the directors fail to take certain measures to monitor and address the company's financial difficulties (reproduced ibid., para. 8). The Swedish company reconstruction order is an insolvency proceeding within the meaning of Art. 2(4) Insolvency Regulation and Annex A to the Regulation. However, the Court of Justice held that the investment company's action did not constitute a closely connected action because it did not concern 'the exclusive prerogative of the liquidator to be exercised in the interests of the general body of creditors' (para. 25). Therefore, jurisdiction was determined by the Judgments Regulation (now the Recast Brussels Regulation), and here specifically by what is now Art. 7(2) Recast Brussels Regulation, which concerns special jurisdiction in matters relating to tort, delict or quasi-delict (para. 42). 54 C-213/10 F-Tex SIA v. Lietuvos-Anglijos UAB 'Jadecloud-Vilma', ECLI:EU:C:2012:215, paras. 40-46, concerning the assignment of an avoidance action. The court in F-Tex acknowledged 'that the right on which the applicant in the main proceedings bases its action is linked with the insolvency of the debtor as it has its origin in the right to have a transaction set aside conferred on the liquidator by the national law applicable to insolvency proceedings' (para. 40). However, it held that 'the exercise by the assignee of the right acquired is not closely connected with the insolvency proceedings […] [since it] is subject to rules other than those applicable in insolvency proceedings ' (paras. 41-42). In the opinion of the court, the claim is different in nature once it has been assigned, because the assignee does not act in the interest of the creditors, but for his personal benefit and can freely decide whether or not to enforce the claim (paras. 43-44). This line of reasoning seems to indicate that the assignee's action is not closely connected because it is not pursued in the interest of the general body of creditors (the second condition of Gourdain v. Nadler). However, if a percentage of the proceeds obtained from enforcing the claim is paid into the insolvent estate, enforcement of the claim is in the interest of the general body of creditors. The main reason, then, for excluding the claim from the scope of the Insolvency Regulation seems to be that it is not brought by the liquidator (the third condition of Gourdain v. Nadler). and private international law if they are enforced by the liquidator, and as company law (or indeed another legal area, for example tort law) if enforced by a third party. This approach may be criticised for rendering the applicable law a function of somewhat arbitrary differences in a Member State's internal laws, which may pursue very similar policy goals with or without the involvement of a liquidator. 55 Nevertheless, it is now firmly embedded in the Court's case law.
Case law furthermore shows that the three conditions apply cumulatively. An action to enforce a claim for payment of services rendered by an insolvent company was held not to be a closely connected action, even though the action of course benefited the general body of creditors by improving the asset position of the company and was, in the case at hand, brought by the liquidator after the opening of insolvency proceedings. The court explained that '[t]he fact that, after the opening of insolvency proceedings against a service provider, the action for payment is taken by the insolvency administrator appointed in the course of those proceedings […] does not substantially amend the nature of the debt relied on which continues to be subject, in terms of the substance of the matter, to the rules of [general commercial] law which remain unchanged.' 56 Conversely, avoidance actions derogate from the rules of general civil and commercial law, but if they are not enforced by the liquidator and instead assigned to a third party, they are no longer 'closely connected with […] insolvency proceedings'. 57 All cases mentioned so far, with the exception of Kornhaas, 58 concerned the scope of the insolvency court's international jurisdiction pursuant to Articles 3 and 6 Insolvency Regulation. Until Kornhaas was decided in December 2015, it was therefore unclear whether the Gourdain v. Nadler criteria also determined the scope of the lex concursus, the applicable law pursuant to Article 7 Insolvency Regulation. 59 In Kornhaas, the Court answered the question in the affirmative. Kornhaas dealt with a reference from the German Federal Court of Justice regarding the characterization of a provision imposing liability on managers of a private limited company incorporated under the laws of England and Wales for payments made after the company had become insolvent. The Court referred to its prior holding in H v. H.K., 60 a case concerning the same provision of German law, where it had held that a national court that has international jurisdiction to open insolvency proceedings pursuant to Article 3 Insolvency Regulation has jurisdiction to rule on such an action. The Court concluded from H v. H.K. that the German provision had to be qualified as insolvency law not only for the purpose of determining the international jurisdiction of the insolvency court, but that it was also 'covered by the law applicable to insolvency proceedings and their effects, within the meaning of Article [7(1)]' of the Insolvency Regulation. 61 The Court added that the German provision fell within the scope of the lex concursus by virtue of Article 7(2) Insolvency Regulation, which provides, inter alia, that the lex concursus shall determine the conditions for the opening of insolvency proceedings. In order to ensure the effectiveness of provisions of national law requiring that insolvency proceedings be opened, the Court argued that Article 7(2) Insolvency Regulation 'must be interpreted as meaning that, first, the preconditions for the opening of insolvency proceedings, second, the rules which designate the persons who are obliged to request the opening of those proceedings and, third, the consequences of an infringement of that obligation fall within its scope.' 62 We will now apply the Court's criteria to those of the above legal mechanisms potentially belonging to either the lex societatis or the lex concursus 63 that have not yet been the subject matter of a decision by the Court (a shift of directors' duties to creditors in the vicinity of insolvency; duty to recapitalise a company; wrongful trading; liability for failure to file for the opening of insolvency proceedings; the liability of directors or shareholders for causing the company's insolvency or frustrating claims of creditors; and the re-characterization of shareholder loans).
First, in obiter, Kornhaas makes it clear that provisions penalising directors for trading at a time when the company should have been put into insolvent liquidation-wrongful trading, liability for failure to file, and similar mechanisms-are principally caught by the lex concursus. 64 However, in spite of the fact that it is universally acknowledged that Articles 6 and 7 Insolvency Regulation have to be interpreted autonomously, Kornhaas does not ensure a consistent treatment of all functionally comparable mechanisms. ÖFAB v. Frank Koot illustrates the problem. The case concerned a provision pursuant to which directors are personally liable for the debts of a company that continues to trade in insolvency. 65 It is therefore functionally equivalent, for example, to wrongful trading under UK law. 66 (2019), Part 1, para. 84. These differences in characterisation are unlikely to disappear completely with a consistent application of the Gourdain v. Nadler conditions, as will be discussed presently in the text. 65 For further details, see n. 53 above. 66 Davies (2006); see also Bachner (2004). individually. The third condition of Gourdain v. Nadler, accordingly, was not satisfied, and the Court of Justice concluded that the action did not fall within the scope of the Insolvency Regulation (however, this does not mean that the provision is to be characterized as company law, as we will see in Sect. 2.2.1 below). 67 Secondly, the liability of a director who does not act in the interest of creditors after duties have shifted from shareholders to creditors in the vicinity of insolvency is typically enforced by the liquidator in the insolvency of a company in the interest of 'the company's creditors as a whole'. 68 Thus, the operation of creditor-regarding duties conforms to the second and third conditions of Gourdain v. Nadler. This leaves the question whether they derogate from ordinary civil and commercial law. Arguably, this question needs to be answered in the affirmative, at least if directors' duties are structured as in the UK Companies Act 2006, which have the interests of the shareholders at their centre, 69 since creditor-regarding duties then deviate from general directors' duties at a particular point in time in a company's life. On this view, liability claims against directors who disregard creditor interests in the vicinity of insolvency would, accordingly, have to be qualified as closely connected actions, provided the liquidator enforces the claims.
Thirdly, the characterisation of the duty to recapitalise a company and of the recharacterization of shareholder loans as equity capital are ambivalent in the Member States, which is not surprising, given the range of designs for these mechanisms in national laws. The duty to call a general meeting and either recapitalise or liquidate a company operates, at least potentially, in the vicinity of insolvency, but it relies on general mechanisms of corporate law to achieve its regulatory goal and, of course, does not involve a liquidator. This may be different, however, where the focus is on liability for the failure to comply with the duty. Here, claims may well be designed in a way resembling liability for a failure to file for the opening of insolvency proceedings or similar mechanisms. 70 The doctrine of equity-replacing shareholder loans can form part of a jurisdiction's capital maintenance regime or be designed as an insolvency law mechanism and entail the subordination of shareholder loans in insolvency. 71 In the former case, where the usual consequences of a violation of capital maintenance rules apply if a shareholder loan that has been re-characterised as equity capital is repaid, the mechanism will typically not derogate from ordinary 67 A provision giving rise to similar classification problems is s. 68 of the Czech Business Corporation Act, see Gerner-Beuerle et al. (2019) (2014), p 326. See also the Italian reform of insolvency law (Legislative Decree of 12 January 2019, no. 14), according to which directors are under a duty to assess whether their company risks becoming insolvent and react promptly to pre-insolvency situations by triggering one of the workout mechanisms detailed in the law (see in particular the new Art. 2086(2) Civil Code). 70 For an example of a provision imposing liability on directors who do not comply with the duty to recapitalise or liquidate a company, see González Fernández (2013) commercial law, whereas it falls outside the scope of the lex societatis in the latter case. 72 The situation outlined above can lead to significant inconsistencies in how the relevant rules are applied where companies exercise their right to choose the applicable company law by relying on Centros. Moreover, the complexity of the questions arising, as well as the large number of host state-home state combinations results in a high degree of legal uncertainty for both companies and their creditors. This is particularly true given that, in many Member States, directors' duties are enforced virtually exclusively in the context of insolvency. 73 As the potential benefits of regulatory arbitrage tend to be modest to start with, it would be unsurprising if costs associated with legal uncertainty and complexity in themselves were sufficient to discourage corporate migration.
Additionally, several of the largest Member States, including Germany, France, Italy, and Spain, rely in important respects on insolvency law, and thus effectively real seat-based rules to regulate corporate behaviour. 74 This is somewhat comparable to the situation in the US, where the scope of company law, according to the internal affairs doctrine, is centred around the shareholder-director relationship, with creditors mainly being protected through insolvency law rules. However, while insolvency law is federal in the US, and thus uniform across all states, it differs significantly across EU Member States. Therefore, for companies primarily operating in a Member State that relies on company law in order to protect creditors, a change in the applicable company law away from this country may result in reduced protection standards, and for companies primarily operating in a Member State that relies on insolvency law to protect creditors, a change in the applicable company law will have a diminished effect on the rules they are subject to, even in circumstances where no duplication of legal requirements results. 75 In contrast, in the US, 72 For example, the German doctrine of equity-replacing loans, in the version in force until 2008, was laid down in a company law statute, Gesetz betreffend die Gesellschaften mit beschränkter Haftung (GmbHG) [Limited Liability Companies Act], ss. 32a, 32b. Section 32a(1) provided that the shareholder's claim for the repayment of the loan was subordinated in the company's insolvency. In spite of its company law provenance, this aspect of the doctrine was characterised as insolvency law, since it only became relevant once insolvency proceedings had been opened, BGH, judgment of 21 July 2011, IX ZR 185/10, BGHZ 190, 364. The German courts had furthermore held that re-characterised shareholder loans were to be treated as equity capital, with the consequence that they could not be repaid while the company was in financial distress (s. 30 GmbHG) and payments made in contravention of the capital maintenance rules had to be returned by the recipient shareholder (s. 31 GmbHG). These aspects of the doctrine of equity-replacing loans were classified as company law for the purposes of private international law, BGH, judgment of 25 June 2001, II ZR 38/99, BGHZ 148,167. The provisions on equityreplacing shareholder loans are now consolidated in the German Insolvency Code, ss. 39, 135, which draw on traditional insolvency law mechanisms to regulate them, notably subordination and the avoidance of repayments within one year before the opening of insolvency proceedings. It is therefore now widely acknowledged that the doctrine is to be characterised as insolvency law, BGH, judgment of 21 July 2011, ibid. 73 See the examples in Gerner-Beuerle and Schuster (2014), pp 326-328. 74 Ibid. 75 On the duplication of rules due to inconsistent characterisation across jurisdictions, see Gerner-Beuerle and Schuster (2014), p 296. See also Mucciarelli (2012), pp 456-457, emphasizing that European company law regimes, in contrast to the US, also tend to embrace creditor protection goals, but that a disagreement about the scope of company law, and more generally the role that company law, insolvency law and other legal areas play in addressing the conflicts that the use of the corporate form gives rise to, is limited, given the common heritage of the 51 US jurisdictions. As a result, the few existing differences in corporate law, for example with regard to power allocation and board insulation, are more relevant when the decision is made where to incorporate, and at the same time it is less costly for firms to incorporate in another state than in the EU, because the problem of over-or underinclusive regulatory regimes, or simply lack of clarity about the applicable rules, does not exist to the same extent. On the other hand, effective protection of creditors based on a company's COMI may render post-Centros opportunities for regulatory arbitrage politically more palatable in the EU.

Effects of Business Activity: Rome II
Conflict rules determining the law applicable to non-contractual obligations generally require a less intense connection with the applicable law than the Insolvency Regulation. In the EU, they are governed by the Rome II Regulation. 76 Company law may overlap in particular with two types of non-contractual obligations, tort law and culpa in contrahendo, in situations where misconduct by a corporate insider gives rise to claims of either the company or third parties against the corporate insider. 77 The connecting factor in tort law is the place where the damage occurs, thus leading to the application of the lex damni, 78 unless the parties involved in the tort have their habitual residence in the same country or the tort is 'manifestly more closely connected' with another country. 79 Culpa in contrahendo is defined as 'a non-contractual obligation arising out of dealings prior to the conclusion of a contract' and is governed by the law that applies, or would have applied, to the contract. 80 Alternatively, if that law cannot be determined, the connecting factors of tort law apply. 81 The lex damni that is applicable pursuant to Article 4(1) Rome II Regulation will, in the present context, generally be the law of the country where shareholders, creditors, or other stakeholders reside who have suffered a loss as a result of the challenged conduct. 82 Misconduct of corporate insiders that amounts to a tortious act within the meaning of Rome II will typically also constitute a breach of directors' duties under company law. Therefore, where claimants are resident in countries other than the incorporation state, the challenged conduct will be subject to at Member States were beginning to relax mechanisms for creditor protection, with the consequence that creditors might increasingly rely on non-company law mechanisms, such as insolvency law. least two simultaneously applying sets of behavioural constraints and sanctioning regimes. This accumulation of legal regimes could be avoided if the law applicable to the tortious act was determined pursuant to Article 4(3) Rome II Regulation instead of Article 4(1), and it could be argued that a manifestly closer connection existed with the country where the registered office of a defendant director's company was located. However, Article 4(3) is described as an 'escape clause', 83 and a general disapplication of the rules of Article 4(1) and (2) in favour of the lex societatis if the defendant is a director or manager of a company is difficult to reconcile with the decision of the drafters of the Rome II Regulation to define the habitual residence of a company as the place of central administration or, when the damage arises in the course of the operation of a branch or other establishment, the place where the branch or establishment is located. 84 The Rome II Regulation defines the term 'non-contractual obligations' negatively by excluding several matters from the scope of the Regulation, in the context of business activities of a company in particular 'non-contractual obligations arising out of the law of companies […] regarding matters such as the creation, by registration or otherwise, legal capacity, internal organisation or winding-up of companies […] [and] the personal liability of officers and members as such for the obligations of the company or body'. 85 The Rome II Regulation therefore confirms that certain questions of core company law, including liability for a breach of directors' duties, are governed by the lex societatis. 86 However, the situation is less clear where the legal mechanism is not directly related to a company's internal governance structure or the position of a defendant as a director or member. In such cases, it is often necessary to determine when legal mechanisms give rise to the personal liability of officers and members as such, and when they impose liability on tortfeasors irrespective of their position as an officer or member. The relevant legal mechanisms can be found in diverse legal areas in the Member States, ranging from company law to tort, quasi-contract, and securities regulation. The situation is further complicated by the use of broadly phrased, open-ended tort law provisions in many Member States that are susceptible to being utilised in a variety of situations closely related to processes within corporations and affecting corporate stakeholders. 87 To give a few examples from national case law, tort law has been relied upon to impose liability on directors for incorrect corporate disclosures 88 or acts that harm creditors. 89 (2), was held to be triggered where a director violated various duties of a criminal and insolvency law nature, including the failure to file for the opening of insolvency proceedings, see BGHZ 126,181. the dissemination of incorrect information to investors and shareholders has been held to constitute a breach of pre-contractual duties (culpa in contrahendo), 90 and in still others a breach of directors' duties under company law. 91 The case law of the Court of Justice offers very limited guidance on the characterisation of the different legal mechanisms, and a common approach to determining the demarcation between the lex societatis and the law applicable to non-contractual obligations has not yet emerged in the EU.
In the following paragraphs, we will discuss the views held in the Member States on this question, focussing on provisions of national law at the intersection of company law and tort law that impose liability on directors, and provisions holding shareholders liable for the debts of a company (piercing the corporate veil). Considering available guidance from the Court of Justice, we argue that liability claims other than those based on a core set of company law duties, for example the liability of directors for a breach of information obligations towards investors or entering into obligations that a director knows the company will not be able to perform, and the liability of shareholders for entering into a transaction that is detrimental to the interests of the creditors and leading to the insolvency of the company, are removed from the scope of the lex societatis.

Liability of Directors
The approaches pursued by Member States in order to distinguish between acts committed by a director that lead to liability under company law and acts that are governed by the law applicable to non-contractual obligations, especially tort law, can be roughly divided into three groups. First, several Member States draw a distinction along the lines of substantive law: liability questions that arise from a breach of directors' duties, the articles of association, or more generally from a breach of company law, are characterised as company law for purposes of private international law, and situations where liability arises from a wrongful act that is not grounded in company law-and that does not consist of a breach of contract or trust either-are characterised as non-contractual in nature and thus subject to the Rome II Regulation. 92 Secondly, in some Member States, conflict of laws characterisation is based on the type of harmful act. If an act involves the exercise of corporate power, it falls within the scope of the lex societatis; otherwise, conflict rules from tort law or another legal area apply. A final approach distinguishes according to the type of injured party: the lex societatis governs any mechanism that gives rise to liability if a loss is caused to the company (and only so-called reflective loss to the shareholders), and the lex loci delicti commissi governs damages claims of third parties that 90 For example, in Germany: BGHZ 71, 284; 72, 382 (dealing with incorrect statements by directors and others to induce investors to invest in a mutual fund or another investment vehicle). More generally see Cartwright and Hesselink (2008). 91  suffer a direct (i.e. not only reflective 93 ) loss. Third parties may include shareholders, if they suffer a loss in an individual capacity, rather than in their capacity as a shareholder because of a reduction in the value of their shareholding, 94 as well as company outsiders, such as creditors or customers.
Of the three approaches, the first one best conforms-with one important caveat that will be discussed presently-to the existing system of European conflict rules as set out in the Rome I 95 and Rome II Regulations. 96 Liability for a breach of directors' duties, giving rise to a claim of the company or against a director, can be interpreted as being a matter of the internal organisation of the company within the meaning of Article 1(2)(d) Rome II Regulation, which is accordingly excluded from the scope of the Rome II Regulation. This view has been confirmed implicitly by the Court of Justice, which decided in Holterman Ferho Exploitatie BV v. Spies von Büllesheim 97 that liability claims based on a breach of directors' duties did not fall within the special tort jurisdiction of the Brussels Regulation. The Court held that where 'a company sues its former manager on the basis of allegedly wrongful conduct, Article 5 (3)  In the case at hand, the action, which had been 'brought by the company against its former manager on the basis of the alleged breach of his obligation to perform his duties properly under company law' was held 'to come within the concept of "matters relating to contract" for the purposes of Article 5(1) of Regulation No. 44/2001' (now Art. 7(1) Recast Brussels Regulation), ibid., para. 54. The action was qualified as contractual because special jurisdiction in matters of contract and tort are regarded as mutually exhaustive in liability actions: 'It is settled case-law that Article 5(3) of Regulation 44/2001 [Art. 7(2) Recast Brussels Regulation] applies to all actions which seek to establish the liability of a defendant and do not concern "matters relating to a contract" within the meaning of Article 5(1) of the regulation [Art. 7(1) Recast Brussels Regulation]', ibid., para. 68. The Court therefore interprets 'matters relating to a contract' as comprising not only the employment or service contract concluded between a director and a company, but generally the legal relationship between them, including statutory duties owed by a director to the company, ibid., para. 69.

Regulations have to be interpreted consistently, 100 the Court's holding is also relevant to a characterisation of liability claims for a breach of directors' duties under Rome II and international company law.
The caveat is that the interpretation of what constitutes a tort/delict cannot be guided by the demarcation between company law and tort law pursuant to a Member State's internal laws, which differ in how they design legal institutions sanctioning misconduct by company directors, but must be based on an autonomous understanding of 'the law of companies' within the meaning of Article 1(2)(d) Rome II Regulation. 101 For example, disclosure obligations that arise when a company accesses public financial markets or is involved in a takeover may be set out in a Member State's internal company law, but it is now widely accepted that liability for incorrect disclosures would nevertheless need to be characterised as falling outside the scope of the lex societatis. 102 The Court of Justice has also held in Harald Kolassa v. Barclays Bank 103 that for purposes of interpreting the Brussels Regulation and determining international jurisdiction, prospectus liability claims as well as damages claims for 'breaches of other legal information obligations towards investors' 104 concern 'matters relating to tort, delict or quasi-delict'. 105 While the connecting factor for jurisdiction ('place where the harmful event occurred') 106 is different from the connecting factor to determine the applicable law pursuant to the Rome 100 Rome II Regulation, Recital 7. 101 In addition, determining the scope of the lex loci delicti commissi by distinguishing between breaches of directors' duties and acts that are not grounded in company law has the disadvantage that it may lead to the cumulative application of two liability regimes if a director's conduct constitutes both a breach of company law and general tort law and the place where the damage occurs pursuant to Art. 4(1) Rome II Regulation is not in the country where the company is registered or incorporated. 102 See Gerner-Beuerle et al. (2019), Part 1, para. 89. However, some uncertainties persist in national law. For example, in the Netherlands it is unclear whether misrepresentations in the annual accounts and reports that cause damage to third parties should be classified as tort law or company law. Vlas (2009), no. 307, suggests that liability is governed by the lex societatis. In Cyprus, prospectus liability is laid down in the Law of Companies, Cap. 113, Art. 43, which has led commentators to conclude that the provision should be classified as company law, Markou and Zantira (2019), Part 2 V, para. 74. In Germany, some commentators suggest a classification of liability for incorrect disclosures to the capital markets as tort law, Assmann and Schütze (2015), § 7, para. 24; Ringe and Hellgardt (2009), pp 809-810, whereas others favour an autonomous classification that relies on the market place where the securities are traded and that has been affected by the disclosure as the relevant connecting factor, Eidenmüller (2004), § 4, para. 36; Grundmann (1990); Hopt (1991), para. 238. 103 C-375/13 Harald Kolassa v. Barclays Bank plc, ECLI:EU:C:2015:37. 104 Ibid., para. 44. 105 Brussels Regulation, Art. 5(3) (now Art. 7(2) Brussels Regulation Recast). 106 The Court of Justice interprets the 'place where the harmful event occurred' as covering 'both the place where the damage occurred and the place of the event giving rise to it, so that the defendant may be sued, at the option of the applicant, in the courts for either of those places', Case C-360/12 Coty Germany v. First Note Perfumes, ECLI:EU:C:2014:1318, para. 46. Where incorrect information is disseminated to the market, the harmful event takes place not necessarily where the investors who have suffered a loss are domiciled, but where 'the decisions regarding the arrangements for the investments […] and the contents of the relevant prospectuses were taken […] or [where the incorrect] prospectuses were originally drafted and distributed', Kolassa (n. 103), para. 53. II Regulation, 107 the underlying policy objectives of both provisions are similar, namely (here) to strengthen the protection of investors in all markets that have been targeted by the issuer of the incorrect statement. 108 Likewise, there are good reasons to conclude that liability for entering into obligations that a director knows the company will not be able to perform is governed by the lex loci delicti. 109 The Court of Justice has not addressed the question directly, but in ÖFAB v. Frank Koot, 110 a case concerning the demarcation between the Insolvency Regulation and the Brussels Regulation, 111 the Court held that 'matters relating to tort, delict or quasi delict' in Article 7(2) of the Recast Brussels Regulation 'must be interpreted as [covering] actions such as those at issue in the main proceedings brought by a creditor of a limited company seeking to hold liable a member of the board of directors of that company and one of its shareholders for the debts of that company, because they allowed that company to continue to carry on business even though it was undercapitalised and was forced to go into liquidation.' 112 The fact that the liability provision at issue was laid down in Swedish company law was irrelevant to the Court's finding that special jurisdiction in matters regarding tort existed. Conversely, where a Member State relies on provisions of general tort law for the regulation of directors' duties, these rules would nevertheless need to be characterised as company law.
In some of the cases outlined in the preceding paragraphs, a characterisation as tort law can have severe consequences. Since the applicable law pursuant to the Rome II Regulation is the lex loci damni and not the lex loci delicti commissi (unless the escape clause of Article 4(3) Rome II can be invoked), directors potentially face liability pursuant to a multitude of ill-aligned legal systems. For example, where the directors continue to trade in violation of legal obligations and creditors enter into contracts with the company, the damage occurs in all countries from which goods 107 Rome II Regulation, Art. 4(1): 'the country in which the damage occurs irrespective of the country in which the event giving rise to the damage occurred', i.e. leading to an application of the lex damni. 108 C-168/02 Rudolf Kronhofer v. Marianne Maier [2004] ECR I-6009, para. 20; Kolassa (n. 103), para. 56 (both dealing with the Brussels Regulation); and recital 16 of the Rome II Regulation. 109 A more problematic case is the liability of directors or other corporate insiders for operating an undercapitalised company that eventually fails, with the consequence that the creditors cannot realise their claims. In some Member States, creditors can bring an action in tort to claim damages, under certain conditions, from the corporate insiders in such a situation. See, for example, the decision of the Dutch Hoge Raad (Supreme Court of Justice) of 6 October 1989, NJ 1990. Creditors were able to rely on the general tort law provision in the Dutch Civil Code, Art. 6:162, where a company had incurred additional obligations at a time when the director knew, or reasonably should have known, that the company would not be able to meet the obligations and the company's assets would not be sufficient to satisfy all claims of the creditors. It has been suggested that liability in such a case should be governed by the lex loci delicti (European Group for Private International Law, Draft Rules on the Law Applicable to Companies and other Bodies, recital to Art. 1(2)(a), reproduced by Garcimartin (2016), p 27), while others argue that the consequences of forming and operating a company without sufficient capitalisation are part of the general rules on capital structure and hence of the lex societatis, Ego (2017), para. 424. 110 See n. 53 above. 111 For the facts, see n. 53. On the relationship between these two instruments, see also n. 32 above. 112 ÖFAB v. Frank Koot (n. 53),para. 42. are delivered or funds are transferred to the company. 113 This seems to run counter to the goal of the Rome II Regulation to 'ensure a reasonable balance between the interests of the person claimed to be liable and the person who has sustained damage'. 114 The second approach, which defines the lex societatis by asking whether an act involves the exercise of corporate power, will in many cases lead to the same result as the first approach. Again, the boundaries between the lex societatis and the lex loci delicti commissi may shift from one Member State to another, since the test depends on how the scope of corporate power is defined in the Member States' company laws. Developing an autonomous definition of 'exercise of corporate power' will be difficult because of the inherent ambivalence of the term and the fact that the extent of corporate power is a function of unharmonised aspects of internal company law. Furthermore, if the criterion was interpreted as implying that directors must have acted within the scope of actual powers conferred on them, it would fall short of capturing some situations that are part of core company law, for example a breach of the duty to act within powers. 115 The third approach encroaches furthest on the rule-making authority of the home state. While a distinction according to the type of injured party (and presumably also according to the type of loss suffered) 116 has the advantage of presenting a relatively clear criterion that allows a functional demarcation between the lex societatis and the lex loci delicti commissi independent of the internal delineation of company law and tort law, 117 it would allow Member States to bring a provision 113 Stone (2018), para. 17.50, with references. The places where the creditor's assets are located or where the creditor is domiciled, on the other hand, are irrelevant, see Kronhofer (n. 108), paras. 19-21. 114 Rome II Regulation, Recital 16. It has also been argued that the rules on international jurisdiction, which often lead to the availability of multiple forums, deter companies from making use of their free movement rights, Dammann (2008), pp 1875-1885. In addition to special jurisdiction in matters of tort law pursuant to Art. 7(2) Brussels Regulation Recast, companies may be sued where they are domiciled, understood as the place where their statutory seat, central administration of principal place of business is located, Arts. 4(1), 63 Brussels Regulation Recast. 115 UK Companies Act, s. 171. 116 Characterising a harmful act as company law or tort law solely on the basis of who has been injured would not always lead to convincing results. For example, it is well established in the Member States that a director who misrepresents facts in disclosures to investors who purchase or sell the company's shares as a consequence of the misrepresentation is liable to the investors under tort law, see, for example, in Germany BGHZ 160,134 (Infomatec I); BGHZ 160, 149 (Infomatec II);andin France Cass. com., 22 November 2005 (Sté Eurodirect marketing c/Pfeiffer), RTD com. 2006, p 445. In order to determine whether a shareholder claim falls within the scope of the lex societatis or the lex loci delicti commissi, it is therefore necessary to rely on additional criteria. In national company law, it is common to distinguish between a loss suffered as a consequence of the invasion of an individual right and a loss that is only a reflection of the loss incurred by the company (reflective loss, see nn. 93-94 above). Shareholders and third parties are in the same position where an individual right of the former has been invaded, and presumably the policy decisions underlying the relevant liability provisions of internal law will take account of the difference in the position of shareholders (and non-shareholders) in such cases on the one hand, and shareholders suffering a reflective loss on the other. Characterisation in private international law should accordingly follow a similar distinction. 117 It also seems to be the preferred solution of the European Group for Private International Law (GEDIP), which has developed draft rules for a regulation harmonising the law applicable to companies in the EU. The draft rules provide that 'liability in tort of the members and directors of a company vis à vis third parties' shall be excluded from the scope of a possible regulation, Art. 1(2)(a) Draft Rules on designed to regulate the behaviour of company directors relatively easily within the reach of host state law (where injured parties are located) 118 by formulating directors' duties broadly and stipulating that the duties are owed not only to the company and shareholders, but also to third parties. The host state could accordingly impose part of its liability regime on the directors of foreign companies operating within its territory in order to protect company outsiders. For example, a formulation of directors' duties as in the French Code de commerce, which provides that directors shall be liable 'to the company or third parties either for infringements of the laws or regulations applicable to public limited companies, or for breaches of the memorandum and articles of association, or for management mistakes' 119 would presumably need to be characterised as tort law according to the third approach in cases where the claimant is a third party. 120 To what extent such a characterisation leads to overreaching host state law depends crucially on the conditions that give rise to liability under national law. Pursuant to the current situation in France, liability towards third parties (understood as not including the shareholders) requires a so-called faute séparable des fonctions (a fault separable from the functions of the defendant director). Faute séparable was described by the Cour de Cassation as 'an intentional fault of a particular gravity that is incompatible with the normal exercise of the director's corporate functions'. 121 This can arguably be equated with a tortious act and may, therefore, be held to justify a characterisation as tort law. However, hypothetically, a Member State may design a liability provision more widely and grant third parties the right to claim compensation for any loss suffered as a result, for example, of negligent management mistakes. Thus, it is clear that this approach to characterisation leads to a potentially broad scope of application of host state law, including in matters that fall within core areas of managerial activity, such as the approval of the company's accounts. 122 In addition, if a third party sues, two or more Footnote 117 (continued) the Law Applicable to Companies and other Bodies. On the proposal, see Garcimartin (2016). A recital to the proposed regulation would clarify that the exclusion applied to liability 'in particular resulting from misrepresentation or undercapitalization', which would instead be governed by the Rome II Regulation, Garcimartin (2016), p 27. Thus, the proposal envisages a bright line rule that includes liability towards the company and shareholders (Art. 5(g) Draft Rules) and excludes liability towards third parties. The rules do not distinguish between direct and indirect (reflective) loss, but shareholders would always be qualified as parties governed by the lex societatis and never as third parties. 118 Art. 4(1) Rome II Regulation. 119 Code de commerce, Art. L225-251 (emphasis added). 120 However, French courts do not follow this approach and characterise Art. L225-251 Code de commerce as company law, irrespective of the claimant, see Cass. civ. 1ère, 1 July 1997 (Africatour), Bull. Joly Sociétés 1997, p 1062, note M. Menjucq (holding that Senegalese law applied to the liability of directors of a Senegalese company towards third parties). 121 Cass. com., 20 May 2003 (Sté d'application de techniques de l'industrie (SATI)), Bull. Joly Sociétés 2003, p 786. 122 Indeed, in more recent case law, the French courts have indicated that an action may constitute a faute séparable where directors exercise their corporate powers, for example to approve financial accounts that are materially misleading, Cass. com., 10 February 2009, appeal no 07-20445 (Société de gestion Pierre Cardin c/Société MMS International). Thus, if characterisation based on the type of injured party became the prevalent approach in the EU to distinguish between the lex societatis and the lex loci delicti commissi, French liability rules in the Code de commerce would have broad application, including in cases where a foreign company operated in France and caused a loss to French residents. liability regimes may apply cumulatively, namely the incorporation state's company law and the tort laws of all countries where the damage occurs, thus possibly resulting in overregulation and overdeterrence.

Liability of Shareholders for Obligations of the Company
In most Member States, the liability of the shareholders for the obligations of the company (piercing the corporate veil) is characterised as part of the lex societatis. 123 However, conceptually, it is not evident why a classification as company law is always the most appropriate solution, and it is indeed possible to find differing views in some Member States and in the academic literature, which suggest a classification as tort law or insolvency law. 124 Considering the criteria established by the Court of Justice to delineate the lex societatis, lex concursus and neighbouring legal areas, it becomes clear that a uniform approach to characterising veil piercing is inadequate, but it is rather necessary to distinguish according to the precise structure, operation and function of individual veil piercing mechanisms that can be found in the Member States. For example, veil piercing according to English law generally applies only in the limited circumstances where 'a person is under an existing legal obligation or liability or subject to an existing legal restriction which he deliberately evades or whose enforcement he deliberately frustrates by interposing a company under his control'. 125 This 123 See Gerner-Beuerle et al. (2019), Part 1, para. 89. 124 See, for example, Calliess (2015), Article 1 Rome II, para. 52 (arguing that piercing the corporate veil should be classified as 'a general problem of (tort) law' and should, therefore, be covered by Rome II). In the Czech Republic, persons (other than directors) who use their influence over the company's directors in a way that results in damage to the company's creditors are liable pursuant to s. 71(3) Business Corporations Act. It is not clear whether this liability should be classified as company law or tort law, but it has been pointed out that it is in character closer to a civil wrong than an obligation under company law, Pauknerová and Brodec (2019), Part 2 VI, paras. 62-63. Controversial is also the characterisation of the German doctrine of causing a company's insolvency (Existenzvernichtung). The basis for the liability of the shareholders pursuant to this doctrine can be found in tort law. In the case law, there is some indication that veil piercing falls within the scope of the lex societatis. In a case dealing both with 'traditional' veil piercing, the liability of the shareholder for the obligations of the company, and in a case dealing with what can be called 'reverse veil piercing', the liability of a company for claims against the sole shareholder, which the shareholder sought to evade by forming the company and transferring assets to that company, the German Federal Court of Justice argued that the liability of the shareholders and the company, respectively, were questions of the reach and meaning of the legal personality of the company (BGH, judgment of 11 July 1957, WM 1957, 1047BGH, judgment of 5 November 1980, BGHZ 78, 318, at III 2 b). They were therefore part of the governing law of the legal person. Whether this jurisprudence can be transposed to the case of liability for causing the company's insolvency is not clear. The liability of the shareholder exists since BGHZ 173, 246 (Trihotel), which introduced a change in the case law, in relation to the company; creditors do not have a direct claim. The doctrine is thus comparable to causes of action of a company against a director who makes payments in the vicinity of insolvency that have been characterised as insolvency law by the Court of Justice (for example, the cause of action discussed in Kornhaas (n. 39). Accordingly, some commentators favour a similar classification of the doctrine of causing a company's insolvency, while others submit that the doctrine is tortious in nature, and yet others that it is part of the lex societatis. An overview of the debate with references is given by Ego (2017) formulation shows that the principle does not constitute a derogation from ordinary civil and commercial law within the meaning of Gourdain v. Nadler. 126 Rather, its aim is to provide a legal response to abuses of the principles of limited liability and separate legal personality generally. 127 However, applying the approaches to distinguishing between the lex societatis and the lex loci delicti described above when we discussed the liability of directors, a characterisation as company law is also by no means clear. Often, veil piercing cases do not involve a breach of company law or an exercise of corporate power, but the disregard or evasion of other applicable rules and regulations, which renders the above criteria ill-suited to identify the boundaries of the lex societatis and lex loci delicti.
In comparison, causing a company's insolvency under German law (Existenzvernichtung), 128 which is commonly described as a case of veil piercing (Durchgriffshaftung), 129 applies to the specific case of shareholders entering into a transaction (or otherwise transferring assets out of the reach of the creditors) in order to benefit certain parties to the detriment of the creditors as a whole and in the knowledge that the action may lead to the company's insolvency. 130 As a consequence, the shareholders are liable to the company for the loss caused by their actions. Given that the company is the claimant, the liability claim will generally be enforced by the liquidator after insolvency proceedings have been opened. Thus, the situation is similar to that of any other legal mechanism imposing liability on directors for acting in a manner that causes a loss to the company's creditors at a time when the directors knew or should have known that their action would cause or aggravate the company's insolvency. The German doctrine should accordingly be classified similarly for purposes of private international law, namely as insolvency law. 131

Outreach Statutes
Where a jurisdiction applies so-called outreach statutes, it does not generally call into question, as a matter of private international law, the applicability of foreign company law rules to an entity incorporated abroad. Instead, outreach statutes can, first, apply specifically to (pseudo-)foreign companies, so as to prevent an unwanted regulatory outcome which would otherwise result from the acceptance of a foreign lex societatis for companies operating in, or interacting with, the economy of the host state. In the context of EU Company Law, the most famous example of such an outreach statute is, of course, the Dutch Law on Formally Foreign Companies, which was at the centre of the Inspire Art case and was held to constitute a restriction on the freedom of establishment that could not be justified. 132 Outreach statutes have also been discussed in the context of regulatory competition in the United States. As examined in another contribution to this special issue, California imposes a number of mandatory corporate law rules on foreign companies with significant economic ties with California, including rules on the election and removal of directors, directors' liability, certain shareholder rights, and recently mandatory gender quotas for directors of public companies. The legality of these attempts to deviate from the internal affairs doctrine is hotly debated in the US literature, and some courts have rejected the applicability of California's corporate law to companies not chartered there. 133 A second, wider category of rules that can be viewed as outreach statutes does not specifically apply to foreign companies, but rather to all entities operating 134 in the host state. Such rules will typically have the same effect as host state rules targeted specifically at (pseudo-)foreign companies where they concern questions addressed at least in part by the company law of the host state, and will thus apply in addition to the lex societatis. Outreach statutes in this second, wider sense typically apply to foreign companies because the legal mechanism they promulgate is formally part of another legal area, for example administrative law in cases where enforcement is through a government agency, or capital markets regulation where the company's securities are listed on a domestic stock exchange. 135 Unless the host state attempts to impose additional requirements only on foreign companies, going beyond the requirements set for companies formed under its own laws, the difference between the two types of outreach statutes is however merely one of legislative technique. 136 Outreach statutes of both types that restrict a company's freedom of establishment may of course be justified under Gebhard, as is, presumably, the case in the fairly widespread practice among Member States of applying domestic director disqualification rules to foreign companies operating within their jurisdiction. 137 Moreover, Kornhaas 138 raises the question whether justification is in fact necessary where the outreach statute in question does not directly affect a company's ability to establish itself in the host state. 139 One may speculate whether the Court is more likely to reach this conclusion in case of the second, wider type of outreach statute. As in the other areas described above, the ability of Member States to impose additional requirements on foreign companies through the use of outreach statutes is likely to reduce the extent to which an exercise of the choice of lex societatis enabled by Centros effectively changes the legal requirements of a company compared to an incorporation in the real seat Member State.

Employees and Defensive Company Law Harmonisation
At least for large companies and corporate groups, the area which would perhaps offer the greatest incentives to engage in regulatory arbitrage in Europe relates to board-level employee participation. 140 Whether affording employees the right to appoint board members (or otherwise 141 influence board composition) creates value overall is of course hard to ascertain empirically. 142 Likewise, the distributional effects of employee participation arrangements for the providers of capital and labour are notoriously hard to measure. 143 Unsurprisingly, therefore, the social optimality of employee participation arrangements continues to be a contested question. 144 Irrespective of the overall ex ante value effects of employee participation, however, shareholders should be expected to potentially benefit significantly from a unilateral right to remove existing protections implicitly provided to employees through employee participation arrangements. As highlighted by Gelter, 145 the ability to disapply employee participation ex post introduces a significant scope for opportunistic behaviour by shareholders and firms. To the extent that employee participation arrangements can be regarded as akin to insurance schemes, with employees accepting lower wages in return for a higher degree of job security, 146 it is obvious that the potential for a unilateral withdrawal of the very mechanism creating the benefit for employees can result in a value transfer from employees to shareholders. From a more dynamic viewpoint, a firm's ability to unilaterally change employee participation rights should in the longer term remove any potential benefits from having these 138 See n. 39 above. 139 See also Sect. 3 below. 140 See on this topic in detail Gelter (2009). 141 For an overview of mechanisms in the Member States see e.g. , Table 2.3.a. France has also subsequently adopted employee participation rules for its largest companies; see Gerner-Beuerle and Schillig (2019), pp 463-464. 142 See e.g. Adams et al. (2010); Davies (2015). 143 See the discussion in , pp 804-805. 144 See e.g. Davies (2015); see also the discussion in Adams et al. (2010) and Kim et al. (2018). 145 Note 140 above. 146 See Kim et al. (2018). arrangements in the first place, whether or not the firm actually engages in opportunistic behaviour. 147 Since the decision in Centros only concerned choice of law at the point of incorporation, it did not directly enable ex post choice of law-and thus opportunistic behaviour of firms vis-à-vis their employees. The decisions that followed, however, confirmed the right of companies to change the applicable law ex post. 148 The two main ways in which ex post choice can be exercised by companies is by way of cross-border mergers and by outright incorporations.
Both of these operations constitute exercises of the freedom of establishment according to the CJEU's jurisprudence, 149 thereby limiting the ways in which the state of emigration can seek to preserve the continued application of its rules due to the strict justification requirements existing under EU law. 150 This may be of particular relevance in the context of employee participation: it is unclear and, it is submitted, doubtful, whether a legal requirement to preserve board-level employee participation, enacted on the Member State level, could pass the Gebhard test. 151 This is partly due to the uncertain nature of the benefits, and indeed the precise aims, of mandatory employee participation rules. 152 The fact that employee participation arrangements can often be sidestepped even within the jurisdictions mandating them 153 casts further doubts on the justifiability of an insistence by a host state on the continued application of its rules following a cross-border reorganisation. Consequently, the choice of law created by the CJEU's case law could have created a significant incentive for regulatory arbitrage, and thus corporate movement, in the area of employee participation.
In reality, however, we observe only very limited regulatory arbitrage around the area of employee participation. There are two principal reasons for the lack of corporate movement away from Member States mandating employee participation.
First, in relation to cross-border mergers, the Member States agreed to 'defensively' harmonise the rules governing this type of transaction in a way that preserved existing employee participation arrangements in the vast majority of cases in which abandoning them could benefit firms or their shareholders. There does not necessarily exist a clear legal basis for concluding that restrictions of fundamental freedoms in secondary EU law can be justified in situations where the same restriction enacted by a Member State could not. 154 However, at least as a matter of fact, the rules protecting employee participation in cross-border mergers, 155 which are based on the SE Model, 156 have not been challenged, and are unlikely to be called into question by the Court. This defensive harmonisation, which had already been enacted by the time the decision in SEVIC Systems 157 was delivered, precluded significant corporate movement motivated by employee participation arbitrage. 158 Second, in relation to cross-border reincorporations, the mere confirmation in VALE and National Grid Indus that such transactions generally fall within the scope of the freedom of establishment ultimately does little to facilitate corporate mobility. Reincorporations are complex transactions, and the CJEU did not, and could not, create a viable legal framework for their implementation. 159 In fact, despite the clear jurisprudence of the Court, several Member States still do not permit reincorporations at all, 160 and even the company at issue in VALE ultimately failed to reincorporate successfully. 161 We are thus likely to only see companies make use of the mobility afforded by these decisions once a directive on cross-border reincorporations-which has already been on and off the EU legislative agenda for decades 162is finally adopted by the Member States. Unsurprisingly, a draft 163 of this directive, which has now cleared most legislative hurdles, adopts essentially the same model for reincorporations as already applies to cross-border mergers and in relation to SE formations.

Limited Review of National Legislative Authority
The preceding analysis, which has identified large areas of law that are removed from the lex societatis and governed by the law of a country or region affected by a company's business activities, and the tentative conclusion that this fragmentation and de-fictionalisation of connecting factors in the EU significantly impedes regulatory competition and corporate mobility, only holds true if deviations from the registered office as the relevant connecting factor and the intervention of host state law are compatible with the free movement rights afforded to companies by the Treaty. In principle, once it is established that a certain activity falls within the scope of a free movement right and a measure of national law is 'liable to hinder or make less attractive' 164 the exercise of the right of establishment, the national measure must be justified by imperative requirements in the public interest. Furthermore, the national measure must be applied in a non-discriminatory manner, it must be suitable to achieve the objective that it pursues and not go beyond what is necessary in order to attain it (Gebhard test). 165 The Gebhard test has proved to be demanding; it requires the host state to show that the imposition of its own law is both sufficient to address the perceived shortcomings of the home state law, and, effectively, that the host state rule in question is optimal in the sense that no other, less intrusive remedy is available. In a complex and interconnected area such as company law, the host state is unlikely to succeed in this endeavour, as demonstrated by the case law discussed above.
In the present context, which concerns the simultaneous application of competing legal regimes as a result of different connecting factors, limited guidance exists on the question of whether and how the Gebhard test applies. The only decision of the Court of Justice that addresses the question at any length, Kornhaas, indicates that a Gebhard justification is not required. The Court distinguished Kornhaas from Überseering and Inspire Art and held that, in contrast to these two cases, the provision of German law at issue in the proceedings 166 concerned 'in no way […] the formation of a company in a given Member State or its subsequent establishment in another Member State, to the extent that [it] […] is applicable only after that company has been formed, in connection with its business, and more specifically, either from the time when it must be considered, pursuant to the national law applicable under Article 4 of Regulation No 1346/2000[now Article 7 Insolvency Regulation 2015, to be insolvent, or from the time when its over- [in]debtedness is recognised in accordance with that national law'. 167 Because of its remoteness from the act of corporate formation or the transfer of a company's seat to another Member State, and thus an immediate cross-border transaction, the Court concluded that the German provision did not affect freedom of establishment.
Some commentators have interpreted Kornhaas as meaning that measures applicable in the vicinity of insolvency do not constitute restrictions on the exercise of freedom of establishment. 168 Others argue, more generally, that duties and liability provisions that find their legal basis in insolvency or tort law, 169 or any creditor protection law that does not directly relate to the company's incorporation, fall outside the scope of freedom of establishment. 170 Going even further, it has been suggested that the right of establishment generally only captures 'corporate matters', 164 Inspire Art (n. 2), para. 133. 165 Case C-55/94 Reinhard Gebhard v. Consiglio dell'Ordine degli Avvocati e Procuratori di Milano [1995] ECR I-4165, para. 37. 166 Section 64 sentence 1 Limited Liability Companies Act, see n. 39 above. 167 Kornhaas (n. 39), para. 28. 168 Armour et al. (2017), p 228. 169 Teichmann and Wolff (2019), pp 254-255. 170 Lindemans (2016), p 877. in particular rules relating to a company's internal governance structure and capital requirements. According to this view, host states could impose any provision on foreign companies that was not classified as company law without coming into conflict with the right of establishment. 171 Yet others regard Kornhaas as a misguided decision that is 'so downright odd that it deserves to be locked into a secure container, plunged into the icy waters of a deep lake and forgotten about'. 172 Kornhaas, it is argued, introduced a form of Keck test by distinguishing between rules of the host Member State that related to the conduct of a company's business, which were outside the scope of the right of establishment, and 'rules that affect the process of setting up the establishment itself', which constituted a restriction that had to be justified pursuant to the Gebhard formula. 173 This distinction, it is further submitted, was difficult to reconcile with both the formulation of Article 49 TFEU and prior case law analysing the rule-making authority of a host state, notably Inspire Art, which seemed to imply that the Treaty's protections were not confined to the act of setting up a company. 174 Rather, the application of any provision of host state law that rendered the exercise of the right of establishment less attractive had to be justified under Gebhard. 175 It is correct that the Keck test has given rise to its own problems of demarcation and the Court has been reluctant to transpose it to areas other than the free movement of goods. 176 Nevertheless, Keck, and similarly Kornhaas, reflect the need to correct the overly broad construction of the scope of the market freedoms in the case law of the Court of Justice beginning with Dassonville. 177 Corrections at the level of justification, while certainly possible, would not be conducive to legal certainty. For example, it could be argued that the application of a legal mechanism characterized as insolvency law for purposes of private international law to a company with its registered office in a different state than its centre of main interest was necessary in order to protect the company's creditors, since the corresponding legal mechanism in the insolvency law of the home Member State (where the registered office of the company is located) did not apply. However, this line of reasoning presupposes that the applicable law of the home Member State (the lex societatis and possibly other legal areas) did not contain a functionally comparable mechanism that protected creditors sufficiently well. Thus, the Gebhard test would require a highly complex (and, arguably, subjective) analysis involving the identification of functional substitutes and an assessment of their comparative effectiveness. Such a view would also suggest that the compatibility of a restrictive national measure with the Treaty would ultimately depend on the specific Member State pairing in question. Furthermore, the assessment would have to change in response to changes to the relevant legal rules in either of the jurisdictions concerned.
Assuming that it is, therefore, preferable to introduce necessary corrections by delimiting the scope of the right of establishment, the next question is how workable criteria can be developed that do not suffer from the same uncertainties that afflict the Keck test. Restricting the right of establishment to provisions 'belonging' to company law (in some sense of the word, for example within the meaning of private international law) is in conflict with well-established case law of the Court of Justice holding that the right of establishment prohibits company law as well as non-company law restrictions, for example exit taxes, and difficult to justify from a policy perspective. 178 Similarly, the view that creditor protection rules are generally outside the scope of the freedom of establishment, provided they do not directly relate to a company's incorporation, 179 is most likely broader than what the Court intended to say in Kornhaas, where it required a certain (albeit badly defined) remoteness between the formation of the company and the point in time when the national measure in question begins to operate. In addition, it would give rise to difficult boundary questions, since many creditor protection rules operate throughout the existence of a company by shaping incentives and deterring particular types of behaviour. 178 BGH NJW 2007, 1529; OLG Rostock, GmbHR 2010, 1349. These cases concerned the liability of directors of foreign limited companies who had acted on behalf of their companies without making sufficiently clear that a legal person with limited liability was one of the contracting parties (notably by failing to use the addition 'ltd.' or a similar designation after the company name). The courts argued that German law was applicable because liability was not based on the position of the director as a corporate organ or the violation of company law duties, but on a quasi-contractual legal mechanism: the creation of the false legal appearance that a person with unlimited liability would be a party to the contract (analogy to s. 179 German Civil Code, a provision of agency law that provides for the liability of an agent who acts without authority). In such a case, the relevant connecting factor is the place where the false legal appearance was created and had an effect on third parties (see BGHZ 43,21,27). Imposing liability on the acting director who failed to use the required designation was not seen as a restriction because the type of liability at issue 'did not fall within the scope of the lex societatis and, hence, did not concern the right of establishment', BGH NJW 2007, 1529, para. 10 (own translation). 179 See n. 170.
The first view mentioned above, excluding measures applicable in the vicinity of insolvency from the scope of the Treaty, 180 is closest to the ruling in Kornhaas, where the Court explicitly referred to insolvency or the vicinity of insolvency as the relevant dividing line. However, as discussed in Sect. 2.1, measures may apply in the vicinity of insolvency that do not constitute a closely connected action and do not form part of the lex concursus as determined by Article 7(2) Insolvency Regulation. If they were excluded from the freedom of establishment, Member States could exploit gaps between the Insolvency Regulation and the right of establishment through an appropriate design of their internal law. A legal mechanism that achieves the desired regulatory outcome may be structured so that it does not fall within the definition of a closely connected action, while the necessary connecting factor is formulated in a way that brings relevant business activity within the reach of host state law. A host state may then be able to apply its laws to a foreign company without the need for justification even in cases where the COMI is not located in that state. It is therefore more convincing to synchronise the international scope of application of the Insolvency Regulation and the scope of the right of establishment. A measure does not per se fall outside the scope of the freedom of establishment if it relates to activity in the vicinity of insolvency, but only if it is characterised as insolvency law pursuant to the Insolvency Regulation. This approach promotes legal certainty, since the international scope of the Insolvency Regulation is now relatively well defined, given the rich body of case law by the Court of Justice that exists on the question, whereas the term 'vicinity of insolvency' is not a well-established concept of EU law and would require further litigation in order to take on a precise meaning. It is also in line with Kornhaas, since the Court, in discussing which measures are excluded from the scope of the right of establishment, refers to companies that 'must be considered, pursuant to the national law applicable under Article 4 of Regulation No 1346/2000[Article 7 Insolvency Regulation 2015, to be insolvent, or [whose over-indebtedness] is recognised in accordance with that national law '. 181 Conceptually, this approach can be understood as the apportionment of spheres of legislative or regulatory authority by the supranational legislator: for some matters (those falling within Article 7 Insolvency Regulation) to the Member State where the COMI is located, and for others (possibly those falling within the lex societatis, or more broadly any measure that determines how attractive an exercise of the right of establishment is, 182 with the exception of those falling within Article 7 Insolvency Regulation) 183 to the Member State where the registered office is located. Freedom of establishment must respect this apportionment of spheres of regulatory authority, since the allocation of rule-making power to the state of the COMI for measures characterised as insolvency law would be largely neutralised if these measures had to be justified under Gebhard. To put the same point slightly differently, it could be argued that the conflict rules assign home state status to Member States for different regulatory spheres. As far as insolvency law (as defined in Article 7 Insolvency Regulation) is concerned, the Member State of the COMI is the home Member State, and as far as company law is concerned (or more generally, for other matters than insolvency law, see the following discussion), the Member State of the registered office is the home state.
If this interpretation is accepted, the question arises whether Kornhaas can be generalised, with the consequence that the regulatory sphere within which a Member State can operate without the need for Gebhard justification is a function of the applicable rules of conflict of laws, or Kornhaas is limited to the interaction between freedom of establishment and insolvency law. In the former case, it would furthermore be necessary to consider whether conflict rules generally have the function of allocating regulatory spheres for purposes of determining the scope of freedom of establishment, or only some conflict rules have this function. For example, at one end of the spectrum of possible interpretations (albeit one that would require a change in the Court's approach to construing freedom of establishment), 184 it might be argued that the right of establishment, insofar as it was exercised by companies, only concerned national measures belonging to 'company law'. The home state (the state under whose laws a company was initially formed) would then have regulatory authority over matters falling within the scope of the lex societatis (this would be the area 'reserved' for home state law, which could define the legal contours of companies as 'creatures of the law' 185 without further scrutiny under freedom of establishment), and host states over all other matters. Of course, there is no reason why the regulatory spheres of the home state and host states should be collectively exhaustive, and what falls outside the scope of home state control must be within the regulatory authority of the host state. Therefore, notwithstanding that it is well established that the home state's reserved area does not go beyond company law, 186 host states may be required to provide Gebhard justification for both company and non-company law measures that make the exercise of the right of establishment less attractive. How far the requirement to justify non-company law matters reaches depends on the question posed above: whether, and which, uniform conflict rules 184 See n. 178 above and the accompanying text. 185 186 Whether it relates only to certain aspects of company law is unclear. The Court of Justice has held that companies cannot rely on Arts. 49 and 54 TFEU to challenge measures of a Member State affecting the conditions required to retain the 'status' as a company incorporated under the law of that Member State, see Daily Mail (n. 185), para. 24; Cartesio (n. 2), para. 110; National Grid Indus (n. 185), paras. 31-33. However, these cases arose in the context of a transfer of seat or place of management, where the legal personality of a company pursuant to the law of the Member State where it was incorporated was the main issue, and hence only limited guidance can be derived from them. exclude areas of regulation, such as insolvency law or tort law, from the scope of freedom of establishment.
In principle, EU regulations harmonising conflict rules, as all other acts by the EU institutions or Member States, have to comply with the Treaty and, therefore, need to be justified if they amount to a restriction on a fundamental freedom. 187 Given the wide formulation of what constitutes a restriction, 188 it can be argued that all conflict rules that rely on a different connecting factor than the registered office and, accordingly, introduce regulatory requirements that apply in addition to a foreign lex societatis to companies established in another Member State, as well as the regulatory requirements applicable pursuant to the conflict rules, render the exercise of the right of establishment less attractive and must pass the Gebhard test. However, as we said above, uniform conflict rules can be understood as apportioning spheres of legislative or regulatory authority by a decision of a supranational rule-maker. Therefore, if a uniform conflict rule is justified under Gebhard, this may be held to apply also to any rule-making activity by a national legislator that falls within the confines of the conflict rule. Such rule-making activity, it could be said, requires no further Gebhard justification, since it is pursued within a 'regulatory sphere' that has been allocated by the EU legislator in compliance with the Treaty.
There are good reasons to assume that the conflict rule of Article 7 Insolvency Regulation does not fail the Gebhard test. While reliance on a connecting factorthe centre of main interest-that is not aligned with the connecting factor that determines the lex societatis may lead to certain friction between company law and insolvency law, because two jurisdictions that may partially overlap or leave regulatory gaps apply to companies with their registered office and COMI in different Member States, 189 the choice of COMI as a connecting factor is unlikely to be manifestly inappropriate or manifestly disproportionate within the meaning of the Court of Justice's case law. First, it should be noted that the connecting factor for purposes of the applicable company law is not harmonised in the EU, and Member States would be able to eliminate any friction between company law and insolvency law, at least for companies incorporated under their own laws, by requiring, either in the form of a conflicts rule as a precondition for an application of their company law or as a requirement of internal company law, that domestic companies must have their real seat within the territory of the Member State. 190 Secondly, the fact that certain measures that may be found in a Member State's company law are excluded from the scope of freedom of establishment under Kornhaas because they are qualified as 'closely connected actions' within the meaning of Article 6(1) Insolvency Regulation should not be interpreted as an encroachment upon the legislative powers of the home Member State in company law matters. The right of establishment is not conceptualised in terms of determining which Member State has authority to regulate in matters of 'company law', but it prohibits any measure that is liable to render the exercise of the right of establishment less attractive. Thus, it is irrelevant that a Member State may regard a measure that satisfies the Gourdain v. Nadler conditions of a closely connected action as belonging to company law, and hence as supposedly falling within the rule-making authority of the home Member State. Rather, the relevant question is whether the conflict rules that can be found in the Insolvency Regulation are designed in a way that is not manifestly inappropriate or disproportionate in achieving the objective of the rules. It is convincing to argue that this is the case, since the Gourdain v. Nadler conditions reflect considerations of procedural efficiency.
Similar considerations apply to Rome II. We have said above that the dividing line between the lex societatis and the lex loci delicti commissi is not well developed at a supranational level, given the lack of case law by the Court of Justice on the matter. However, it is not manifestly inappropriate for an autonomous definition of the scope of the lex loci delicti commissi to capture provisions imposing liability on directors in cases where the substance of the matter is not related to directors' duties under company law (even though a violation of applicable laws may also amount to a breach of directors' duties), which is the essence of the first approach to distinguishing the lex societatis from the lex loci delicti commissi that we described above. 191 Examples are incorrect disclosures in financial markets 192 and the misconduct of directors giving rise to liability in an action that would have been qualified as a closely connected action had it been brought by a liquidator 193 (that is, the first of the Gourdain v. Nadler conditions-a derogation from the common rules of civil and commercial law-is satisfied, but not the third). In these cases, policy considerations apply that are different from those underpinning general directors' duties, notably the integrity of financial markets, the protection of investors in public markets, and the avoidance of risk shifting in the vicinity of insolvency, which justify a treatment of the respective liability actions in the same manner as other actions where these policy considerations are at the forefront.
In summary, the solution suggested here would result in a layered review of provisions that have the potential to dissuade parties from exercising their right of establishment. First, provisions of national law adopted pursuant to a uniform conflicts rule that allocates regulatory authority to the respective Member State in compliance with the Treaty would not require justification (this is the Kornhaas scenario). Secondly, uniform conflict rules laid down in measures of EU law require justification, but with 'low intensity review' based on the criteria of manifest inappropriateness or disproportionality. 194 Thirdly, outreach statutes, including any explicit or hidden conflict rule in national law that allows a Member State to apply the outreach statute to a foreign company, are subject to a full Gebhard review.

Conclusion
How much choice did the Centros line of cases actually provide to companies and entrepreneurs in Europe? In this article, we try to show that many of the rules that govern a company's activities are not reliably determined according to the lex societatis across the EU. Not only do significant uncertainties exist with regard to the scope of the lex societatis, host states retain considerable power to interfere with the use of foreign company law concepts within their jurisdiction. In addition, Member States often have a choice between a range of functionally equivalent (or at least similar) legislative techniques to achieve a desired regulatory outcome. Where-as in the case of employee participation-significant scope for regulatory arbitrage may exist, Member States have chosen to remove the incentives for engaging in arbitrage through defensive harmonisation.
As a result, choosing what can reliably be considered to constitute company law in isolation, while remaining subject to a plethora of rules from adjacent and competing areas of law, is likely to offer only very limited advantages to companies. This is particularly true for companies having their real seats in jurisdictions that rely extensively on insolvency and tort law to regulate corporate behaviour, as is the case for most of the largest EU economies. On the other hand, choosing to be governed by a foreign company law almost always increases the complexity and legal uncertainty surrounding a company's legal requirements. Outside of the much discussed, but ultimately arguably less important areas of minimum capital requirements and incorporation formalities for micro companies, these costs largely seem to have outweighed the modest benefits of choosing between Member States' company laws. As there is little reason to believe that this situation will change in the foreseeable future, EU company law scholars may continue to be the group most affected by Centros and its progeny.