Defining groups of companies under the European Insolvency Regulation (recast): On the scope of
 EU
 group insolvency law

[Correction added on 28 December 2020, The placement of charts were previously incorrect and have been updated in this version] Abstract This article examines the general scope of application of the provisions on insolvency regarding members of a “group of companies”, as included in Chapter V of the European Insolvency Regulation (recast) (“Recast EIR”), in order to review whether that scope is appropriate to deal with the different group structures in which business may be conducted. With the definition for a “group of companies” playing a paramount role in determining the scope of these provisions, the article includes a thorough analysis of the current definition for a “group of companies” as included in the Recast EIR. Based upon a teleological approach, the article argues in favour of an independent, broad and flexible interpretation of “group of companies”, in order to include a large number of groups within the scope of the Recast EIR's provisions regarding group insolvencies.


| GROUPS OF COMPANIES AND THEIR SPECIFICS IN INSOLVENCY
Over the last century, business enterprises have increasingly organized themselves as multinational groups of companies. Such groups have become "the prevailing form of European largesized enterprises". 1 It has likewise become increasingly important that insolvency laws are able to deal with the challenges specific to cross-border insolvencies involving groups of companies.
interests ("COMI") of the insolvent group members are not (all) located in the same Member State. Group members participate in a GCP on a voluntary basis.
In comparison with its predecessor, the Recast EIR undoubtedly allows for a more coordinated approach to cross-border insolvency proceedings involving groups. Chapter V's new provisions have, however, been met with scepticism. 11 Cross-border communication, cooperation and coordination between insolvency practitioners and courts may be insufficient, because such activities do not necessarily prevent fragmentation of the group, disturbance of the chain of command and a domino effect, even if a GCP is opened.

| From big to small and intricate to simple
Groups of companies manifest themselves in all shapes and sizes. There are, for example, widely known, colossal, worldwide-operating groups of companies, such as Lehman Brothers, which, at the time of its bankruptcy filing, was estimated to have been comprised of approximately 8,000 corporate entities in 40 countries. 12 There are also groups of companies consisting of merely a natural person and two companies of which he or she is sole shareholder, or of a parent company and a single subsidiary. Commonly used group structures may vary among states as well because of differences in legislation. 13 Although groups of companies are comprised of separate legal units, they coordinate their conduct to a certain degree on a structural basis. There is some form of "group discipline" enabling such coordinated conduct.

| Subordination groups
When discussing groups of companies, the prototype majority equity-based group structure first comes to mind: a parent company holds (at least a majority of) the shares in one or more direct or indirect subsidiaries, with a group structure often shaped in a pyramid form. 14 The capacity to impose group discipline and coordinated management is generally based on ownership 15 : the group discipline is premised on the "trickle-down" powers of shareholders based on their voting rights, which allow the parent company to impose its will on the direct or indirect subsidiaries. If the directors of an indirect subsidiary are, for instance, unwilling to follow directions from the parent company, the shareholder will be able to remove the sitting directors and appoint new ones that are willing to follow his directions.
However, the ability to impose group discipline and, thus, a coordinated group policy, does not necessarily require a position as majority shareholder: a minority shareholder may also be able to control a subsidiary by means of specific voting rights attached to his or her shares (whether or not pursuant to the subsidiary's articles of association or bylaws) or an agreement with the other shareholder(s) and/or the subsidiary itself. 16 Group discipline may also be based entirely on agreements between the group companies, rather than (partial) ownership. Franchise relationships have been referred to as an example of such a contractual group relationship. 17 These structures often occur in businesses where brand loyalty is an important factor in consumers' choices. The franchisees are allowed to use the company's brand, its concept and back-office facilities in exchange for the payment of a certain fees. In such structures, the franchisor exerts a certain degree of dominance over the franchisee(s) pursuant to the franchise agreement to ensure uniformity. As a result of this uniform presentation, franchise companies will often be perceived as part of a single business enterprise by the general public, which is likely not able to distinguish between group-owned stores and franchise stores.

| Heterarchical groups
Whether more or less common, the above-mentioned types of group structures all build on a form of subordination to ensure group discipline. Group structures can also be based on a heterarchical (non-hierarchical) relationship, as is exemplified by a type of group known under German law as a Gleichordnungskonzern. Such groups are comprised of legally independent companies, without one company being dependent on the other, while all of them are managed on a unified basis (unter einheitlicher Leitung zusammengefaßt). 18 Group discipline can be maintained by, among other things, appointing the same persons as board members for multiple companies, a management structure that is common in family businesses. 19 Groups can also be a combination of heterarchical and subordinated group companies, as is exemplified by the Anglo-Dutch consumer goods company Unilever. 20 This corporation has two holding companies, a UK Public Limited Company (PLC) and a Dutch Naamloze Vennootschap (NV). 21 Both holding companies jointly and independently own shares in operating companies (Chart 1). 22 Group discipline is maintained by means of several agreements between the two holding companies (the so-called "Foundation Agreements") and by implementation of specific provisions in the holding companies' articles of association. Both holding companies have the same directors and their shareholders receive dividends on an equalized basis. 23 The famous Eurotunnel group had a similar heterarchical holding structure, with two holding companies: one at each end of the tunnel the group owned and operated (SA Eurotunnel in France and Eurotunnel plc in the UK). 24 C H A R T 1 Unilever structure. Source: Author's own

| Integration and centralization
Apart from the legal structure used to impose group discipline, the degree to which groups of companies function as an integrated and interdependent business unit may also vary widely. 25 The insolvency law remedies that the group requires may differ depending on the degree of integration and interdependency within the group: a group that was highly integrated before the opening of insolvency proceedings would have more need for coordinated conduct of such proceedings than several loosely connected companies.
Mevorach has identified three levels of integration and interdependency relevant to cases of group insolvencies, 26 and mentions several factors that may be relevant in assessing these levels, amongst which are the following: 1. the level of economic integration (the extent to which companies engage in a common business); 2. the degree of actual control exercised by a central management over group companies; 3. the extent to which group companies rely on other group companies for vital functions (e.g., legal services, financing, accounting, warehousing); 4. the existence of group financing (e.g., the existence of one or multiple lending facilities on a group level for which some or all of the group companies are jointly and severally liable or the existence of group guarantees); 5. the extent to which personnel is rotated for assignments amongst group companies; and 6. how the group presented itself to the public, and to what extent it was perceived by the public as a single entity. 27 In some cases, groups of companies are merely loosely connected, conducting segregated businesses that operate on an autonomous basis, without any group discipline being exercised. It is also possible that, although part of a group runs an integrated business, other parts of the group may form independent units. For example, this is typically the case with SPVs (special purpose vehicles) or companies held for investment purposes. Mevorach has defined this level of integration as "weak or non-integration". 28 However, a group of companies will often contain multiple integrated and interdependent companies, all of which are focused on a part of a single business enterprise, with, for instance, the same directors throughout the structure, group companies being dependent on the group for financing, joint brand management and a joint strategy determined by the parent company. 29 It is precisely in these groups of companies that the opening of insolvency proceedings may lead to a domino effect: because of the interdependency of companies within the group, the insolvency of one or several group companies generally leads to insolvency proceedings being opened throughout the group. 30 Surplus value in the group as a whole is also very likely to exist in these groups. Mevorach identifies this level of integration as "business integration", which is understood to mean a "business [which] was operating with significant integration or inter-dependence, so that the group was unified in terms of its business". 31 Integration and interdependency can go even a step further, in the sense that assets and/or debt of group companies are intermingled, for example, as a result of negligence or bad faith. Mevorach refers to this third level of integration as "asset integration". 32 Asset integration is unlikely to occur often in the case of multinational groups of companies. National laws will generally require companies to maintain their books and records on an entity-by-entity basis. 33 With the financial reporting of larger-size companies generally being subject to audit by external controllers, groups of companies will generally conduct their business in a manner that prevents extensive intermingling of assets and/or debts.
In addition, besides varying in degree of integration and interdependence, groups of companies may also differ in the degree to which a group is centrally controlled. Management within groups of companies can be highly centralized and conducted from a group-wide head office, or can take place with a significant degree of decentralization. 34 Although an integrated group of companies is likely to be centrally managed via a head office, there may be cases in which subsidiaries of an integrated group of companies are managed in a decentralized manner and granted significant autonomy in determining their policy and strategy, or even cases in which the group lacks a centralized head office and is managed on a heterarchical (non-hierarchical) basis. 35 It is necessary to review to what extent all of the above-mentioned groups fall within the scope of Chapter V's provisions. Where that is the case, insolvency practitioners and courts in insolvency proceedings involving such group members may be subject to group-specific obligations under the Recast EIR and have access to legal coordination instruments to promote efficient management of those proceedings, which instruments would otherwise not be available.
The most important limitation to Chapter V's application is found in the Recast EIR's definition of groups. In general, the definition of a group of companies has differed depending on the objectives of the provisions in which the definition appeared, such as accounting law, tax law and laws governing corporate liability within groups of companies. 36 Member States' national laws have longer included definitions of group companies, 37 and the EU has already been confronted with the challenge of devising a definition of the term in connection with, among other things, drafting the directive on (consolidated) financial statements ("Accounting Directive"). 38 When drafting the Recast EIR, there was, however, no consensus on the definition that should be used in the context of international insolvency law. 39 One of the main challenges in defining groups of companies in the insolvency context is that the definition should provide sufficient certainty and predictability, while at the same time be flexible enough to include all relevant types of current and future group structures. 40 On the one hand, it is clearly beneficial to a market when business participants are able to assess ex ante their risks in contracting with other business participants, including their risks in case of insolvency of their counterparty. 41 Legal measures that are aimed at treatment of the group's insolvency as a whole can substantially impact the position of creditors, as will be further discussed below. Counterparties of a group company should thus be able to predict, with a clear degree of certainty, what their rights and claim(s) would be in case one or more of the group's members enter into insolvency proceedings. In determining the scope of application of insolvency remedies aimed at groups of companies, the definition for "groups of companies" generally plays a paramount role and should allow counterparties of the group companies to predict whether the group of companies would be treated as such in case of an insolvency proceeding.
The more intrusive the group insolvency remedies in a certain legal system are, the greater the need for a predictable definition for "groups of companies". Any definition for "groups of companies" should however, on the other hand, be flexible enough to include all relevant types of current and future group structures. A formal standard, such as a specified minimum percentage of ownership or voting rights in a subsidiary, would provide certainty, but may be prone to misuse, might be inflexible and would often exclude from its scope what generally is considered a group of companies. By contrast, a more open-ended definition, such as "(the ability to) control" would provide for more flexibility, but would provide significantly less certainty. 42 The Recast EIR defines a "group of companies" as a "parent undertaking" and all its subsidiary undertakings. 43 A parent undertaking, in turn, is defined as "an undertaking which controls, either directly or indirectly, one or more subsidiary undertakings. An undertaking which prepares consolidated financial statements in accordance with [the Accounting Directive] shall be deemed to be a parent undertaking". 44

(author's amendment)
Although the Recast EIR does not include a definition of "subsidiary undertaking", an inference from the definition for "parent undertaking" suggests that a subsidiary undertaking is an "undertaking controlled, either directly or indirectly, by another undertaking" (i.e., the parent). 45

| Control as a primary defining concept -But what is it?
The Recast EIR's definitions use the term "control" as the defining characteristic for groups of companies. Understanding this concept is therefore paramount to understanding what constitutes a group within the regulation's framework. When the European Commission presented its proposal for the amendment of the Original EIR in December 2012 ("Commission Proposal 2012"), it largely maintained a formal approach towards "control". It limited the concept to a large extent to the "trickle-down" powers of shareholders that follow from their ownership position. 46 According to the definition of "parent company" in the Commission Proposal 2012, control would exist if a company either had the majority of shareholders' voting rights in a subsidiary company or, in the absence of such a majority, was a shareholder with the right to 1. appoint or remove a majority of the members of the administrative, management or supervisory body of that subsidiary; or 2. exercise a dominant influence over the subsidiary company pursuant to a contract entered into with that subsidiary or a provision in its articles of association. 47 In its response to the Commission Proposal 2012, the European Parliament ("EP") proposed amending the definitions for groups to the more abstract and open form currently used in the Recast EIR. 48 This was done in order to align the Recast EIR with the definitions for groups in the Accounting Directive, 49 which includes almost identical definitions to those contained in the Recast EIR. 50 Contrary to the Commission Proposal 2012, the Accounting Directive applies a more open approach toward "control". While neither the Accounting Directive nor the multiple documents resulting from its legislative process provide a direct definition of "control", 51 its recitals stipulate that "Undertakings controlled by the parent undertaking should be considered as subsidiary undertakings. Control should be based on holding a majority of voting rights, but control may also exist where there are agreements with fellow shareholders or members. In certain circumstances control may be effectively exercised where the parent holds a minority or none of the shares in the subsidiary.
[…]" 52 Although starting with a reference to the majority of voting rights as a basis for control, the Accounting Directive recognizes that control may also be effectively exercised where the parent holds a minority or none of the shares in the subsidiary.
With the goal of more closely aligning the definitions under the Recast EIR and the Accounting Directive, and observing that both definitions are nearly identical, it has been argued that the definitions relating to "groups of companies" included in the Recast EIR should be interpreted in accordance with the Accounting Directive as far as possible. 53 This has been stated to have "the great benefit of systematic consistency and coherency in EU law". 54 However, although consistency and coherency are valuable objectives, aligning the Recast EIR's definitions with those in the Accounting Directive may not be as advisable as it appears at first blush.
Firstly, the Accounting Directive is a directive that requires implementation in Member States' national laws and leaves considerable room for variation. 55 The Dutch implementation of the Accounting Directive and its predecessor (the Seventh Council Directive) is a good example of such variation. The current definition of "groups" in the Dutch Burgerlijk Wetboek (the Dutch Civil Code), which was introduced in response to the Seventh Council Directive, 56 does not prescribe any form of control, contrary to the directive. 57 Instead, its group concept is based on "economic unity" and "organizational linkage". The Dutch definition of subsidiaries (dochtermaatschappijen) includes no reference to "control" and a remarkably narrow description of the parent-subsidiary relationship. 58 The obligations governing preparation of consolidated financial statements were actually included in a separate provision and do not necessarily require qualification as a "group". As a result, the Dutch implementation of the Accounting Directive has been described as "somewhat quirky". 59 Moreover, as will be further discussed below, 60 Member States were intentionally given considerable discretion in implementing the Accounting Directive. With such differences in Member State's implementations, it is questionable whether adhering to the Accounting Directive would actually promote consistency and coherency in interpreting the definition of "group of companies".
Secondly and on a similar note, while the Accounting Directive requires implementation and may differ from one Member State to another, the Recast EIR has direct effect in all Member States. 61 As stressed by the Court of Justice of the European Union ("CJEU") when ruling on the interpretation of COMI under the Original EIR "The [CJEU] has consistently held that it follows from the need for uniform application of European Union law and from the principle of equality that the terms of a provision of that law which makes no express reference to the law of the Member States for the purpose of determining its meaning and scope must normally be given an autonomous and uniform interpretation throughout the Union, having regard to the context of the provision and the objective pursued by the legislation in question…" 62 (author's amendment) The EU's "group insolvency law" tries to solve the challenge of efficiently managing insolvency proceedings regarding group companies and leveraging potential synergy value of the group as a whole, while simultaneously respecting that group companies constitute separate legal entities. 63 Those objectives, set within a more elaborate set of objectives specific to the Recast EIR, differ significantly from the Accounting Directive's objectives of comparability and equivalence in information that undertakings publicly report and to ensure that annual financial statements give a "true and fair view of an undertaking's assets and liabilities, financial position and profit or loss". 64 The definitions of "group of companies" and "parent undertaking" in the Recast EIRand thus the concept of "control" in Article 2 (14), 1st sentence, of the Recast EIRshould thus have an autonomous interpretation in light of the Recast EIR's own context and objectives, rather than simply adhering to the Accounting Directive. 65 In light of Chapter V's specific objectives, the Recast EIR's provisions on group membersincluding the definitions in Articles 2 (13) and (14) should be interpreted in a way that would allow for efficient management of the relevant insolvency proceedings, as long as the separateness of the group companies and the Recast EIR's more general objectives are not jeopardized.
For purposes of interpreting "control", the work of UNCITRAL on cross-border group insolvencies provides a valuable source for comparison. 66 UNCITRAL recently developed its Model Law on Groups, a new model law intended to apply to cross-border insolvencies for groups of companies. 67 The "Enterprise Group" is the Model Law on Groups' equivalent of "group of companies" as used in the Recast EIR. An "Enterprise Group" is defined as "two or more enterprises that are interconnected by control or significant ownership" , 68 with control being "the capacity to determine, directly or indirectly, the operating and financial policies of an enterprise". 69 Thus, UNCITRAL has explicitly chosen a more open approach towards control, instead of a formal standard.
UNCITRAL describes the key elements of "control" to include "actual control or capacity to control, either directly or indirectly, financial and operating policy and decision-making. Where the definition includes capacity to control, it generally envisages a passive potential for control, rather than focusing upon control that is actively exercised. Control may be obtained by ownership of assets, or through rights or contracts that give the controlling party the capacity to control. What is important is not so much the strict legal form of the relationship, between the entities, such as parent and controlled group member, but rather the substance of that relationship". 70 UNCITRAL's work represents internationally accepted standards for international insolvency law. Contrary to the Accounting Directive, the Model Law on Groups to a large extent adheres to the same objectives as Chapter V. 71 In the absence of any indication of an interpretation of "control" in the Recast EIR itself, "control" as used in Article 2 (14), first sentence, of the Recast EIR should be interpreted in conformity with the definition used in the Model Law on Groups and UNCITRAL's preceding work on groups of companies, 72 namely "the capacity to determine the operating and financial policies of an undertaking". 73 Whether or not an undertaking has the capacity to "control" another undertakingand thus qualifies as a "parent undertaking"should be determined by assessing all relevant factors. Factors that may indicate the existence of control, according to UNCITRAL, include, inter alia, the ability of an entity 74 1. to dominate the composition of the board of directors or governing body of a second entity; 2. to appoint or remove all or a majority of the directors or governing members of a second entity; 3. to control the majority of the votes cast at a meeting of the board or governing body of a second entity; and 4. to cast or regulate the casting of a majority of the votes that are likely to be cast at a general meeting of a second entity, irrespective of whether that capacity arises through shares or options.
Other relevant factors that could be considered in determining whether "control" within the meaning of Article 2(14) of the Recast EIR exists are, for example 1. whether particular rights are awarded to an entity in a second entity's articles of association, or in agreements between those parties, that limit the capacity of the second entity's board of directors to determine its operating and financial policies 75 ; 2. whether voting rights and rights of appointment and removal attached to shares in a second entity are exercised by another entity than its shareholder(s) (e.g., on the basis of a right of pledge) 76 ; and 3. whether shares in an entity are held on behalf of another entity than the potential parent or a subsidiary of that potential parent. 77 Therefore, typical rights given to (majority) shareholders, such as the ability to appoint and remove a majority of the directors of a company and the right to determine the majority of votes cast at the general meeting of a company, are factors that may constitute "control". From a teleological perspective, however, it would however not be advisable to limit "control" to equity ownership. Chapter V's provisions are aimed at ensuring the efficient administration of insolvency proceedings involving different group members, while respecting each group member's separate legal personality.
In keeping with those objectives, the legal form of the relationship should not be decisive. Rather, it is vital that a certain degree of integration and/or central management has existed prior to insolvency, resulting in synergy value and/or cost efficiency if the insolvency proceedings are managed on a coordinated basis, to the extent that does not harm the group members' legal separateness. Moreover, the Recast EIR itself nowhere implies that "control" could only be based on ownership of one undertaking by another. Additional support for this argument can be found in the Commission Proposal 2012, which originally explicitly limited "control" to situations in which a parent was a shareholder of its subsidiary, but was amended in 2014 to remove the reference to equity ownership from the definition of a "group of companies". 78 A contractual network of franchise companies and other similar contractual networks may also constitute what could be considered a group of companies. 79 Himmer has argued that, in general, franchise networks constitute too loose a connection to be deemed a "group of companies" within the meaning of Article 2(13) of the Recast EIR. 80 According to Himmer, any communication, cooperation and coordination in insolvency proceedings involving franchise companies should take place on a consensual basis, outside the framework of Chapter V. As was argued above, however, the legal form of the relationship is not decisive in determining whether "control" exists in relation to two "undertakings".
Thus, it should not be decisive that the relevant relationship is governed by a "Franchise Agreement". Rather, the actual content of the relationship between the companies is decisive for determining whether control exists. In general, it could be argued that franchisees and their creditors would often benefit from group-wide insolvency solutions, as they are often to a large extent integrated within the group of companies, for example, by virtue of the same branding, as they sell the same products or services and as they generally make us of the group's backoffice functions (e.g., a joint purchasing and logistics network). Depending on the circumstances at hand, qualifying franchisees as group companies could thus potentially benefit the Recast EIR's goals of efficient management and leveraging synergy value.
The same broad approach applies to the question of whether an undertaking must actively exercise its control to qualify as a "parent undertaking" under the Recast EIR. The potential that a parent undertaking may exercise control (e.g., by removing and appointing a majority of the board members) will generally be sufficient to ensure group discipline, without that parent actually exercising its control. 81 Note, however, that pure economic power over, or the economic dependency of, another undertaking should be insufficient to constitute "control", in itself. Otherwise, the market dominance of certain market participants, for example, monopolists or oligopolists, would generally be sufficient to constitute a parent-subsidiary relationship within the context of article 2(13) of the Recast EIR.
A broad and flexible interpretation of "control" as argued above may inherently have a negative impact on the ex ante ascertainability for third parties whether a company is part of a "group of companies" within the framework of the Recast EIR. However, Chapter V's provisions concerning "groups of companies" in the Recast EIR have a very limited impact on the relevant stakeholders in the individual insolvency proceedings: the GCP is based on voluntary participation, and the duties which Chapter V's Section I imposes on insolvency practitioners and courts are, among other things, limited to cases where such cooperation would actually be beneficial to the relevant insolvency proceedings. 82 As such, this broad and flexible approach should not be problematic. This reference is significant because, as a result of it, the interpretation of the definitions for "parent undertaking", and thus "group of companies", will differ among Member States, as discussed in more detail below.
In implementing the Accounting Directive, Member States were required to obligate parent undertakings to prepare consolidated financial statements, which also include those of their subsidiary undertakings, in relation to a parent undertaking that 83 1. holds the majority of shareholders' voting rights in a subsidiary undertaking; or 2. in the absence of such majority is, among other things, a shareholder with the right to: a. appoint or remove a majority of the members of the administrative, management or supervisory body of the subsidiary; or b. exercise a dominant influence over the subsidiary company pursuant to an agreement entered into with the subsidiary or in accordance with a provision in its articles of association.
Apart from these requirements, Member States were granted significant discretion in implementing the Accounting Directive. Member States could, for example, exempt mediumsized groups from the obligation to prepare consolidated financial statements (while small-sized companies were exempted entirely). 84 Member States could also elect to apply the obligation to prepare consolidated financial statements to 1. parent undertakings that do not own shares in their subsidiary undertakings but nevertheless (have the power to) exercise a dominant influence or control over them 85 ; or 2. undertakings that are not considered part of a "group" as they lack any form of "control", but are managed on a unified basis or because a majority of the persons making up their administrative, management or supervisory bodies are the same (e.g., the German Gleichordnungskonzern). 86 As a result, the interpretation of "group of companies" under the Recast EIR may differ in each Member State because 1. the preparation of consolidated financial statements is a sufficient condition to qualify as a "parent undertaking"; and 2. the types of groups that are obligated to prepare consolidated financial statements may differ in each Member State.
Although this may result in an inconsistent interpretation of "group of companies" under the Recast EIR, this should not necessarily pose a problem. Companies included in consolidated financial statements will generally be viewed by creditors and other stakeholders as together forming a single group. The companies themselves have created the legitimized expectation that they are doing business as a group. If companies prepare consolidated financial statements, they are also likely to be integrated to a certain extent, supporting the case for a form of coordination between their insolvency proceedings.
Whether or not the presumption in Article 2 (14), 2nd sentence, of the Recast EIR should be rebuttable has been a subject of debate. Some argue that a rebuttable presumption would allow for a flexible interpretation taking insolvency-specific objectives into account when construing the definitions related to groups of companies. 87 The wording of Article 2(14), 2nd sentence, of the Recast EIR, however, leaves little room for discussion: the definition states that such undertakings "shall be deemed" parent undertakings, without any further qualification.
By comparison, Article 3(1) of the Recast EIR also includes a presumption, but explicitly provides that the presumption is rebuttable with evidence to the contrary. 88 The aim of the presumption in Article 2(14), 2nd sentence, of the Recast EIR appears to be to provide a level of ease and certainty in determining whether a group qualifies as such under the Recast EIR, as the general norm of "control" is less precise. Making the presumption rebuttable would detract from these objectives. An undertaking that prepares consolidated financial statements pursuant to the Accounting Directive should automatically qualify as a parent undertaking without further inquiry into the existence of any "control". 89 Qualification as a "parent undertaking" would be warranted in such cases because, as noted previously, the undertaking represented itself to third parties as the head of a group of companies, and there would undoubtedly be some degree of integration.
The same should apply to undertakings that voluntarily prepare consolidated financial statements in accordance with the Accounting Directive but are not obligated to do so under the legislation implementing the directive in their Member State. Qualification as a "parent undertaking" is warranted from a teleological perspective. 90 The wording of the presumption supports this view: "An undertaking which prepares consolidated financial statements in accordance with [the Accounting Directive]".
If the presumption was intended not to apply in such cases, the provision should have stated that "An undertaking which is required to prepare consolidated financial statements…". 91 Although Article 2(14), 2nd sentence, of the Recast EIR does not stipulate so explicitly, it follows from the presumption that the other undertakings included in such consolidated financial statements prepared by the "parent undertaking" should be presumed to be subsidiary undertakings and, thus, part of the relevant "group of companies". 92 The presumption however leads to a peculiar result with respect to heterarchical groups. The definition for "groups of companies" assumes a single ultimate parent undertaking at the top of the group. Heterarchical groups however lack such a single parent undertaking. Within the context of the Accounting Directive, this was solved by stipulating that, in case of heterarchical groups of companies, the relevant references to "parent undertaking" shall be understood to refer to all relevant undertakings. 93 The Recast EIR lacks a similar provision.
However, the Recast EIR expresses a clear intention for a non-rebuttable presumption that companies that prepare consolidated financial statements form a group with the other companies included in those financial statements. Any other interpretation of Article 2(14), 2nd sentence, of the Recast EIR would render this presumption superfluous when read together with the general rule included in the first sentence of the provision. Accordingly, it has been proposed that Article 2(13) of the Recast EIR should be read to include all members of a heterarchical group that are included in the relevant consolidated financial statements. 94 The Recast EIR does not specify when the consolidated financial statements must be or have been prepared in order to qualify for the presumption. Article 2(14), 2nd sentence, of the Recast EIR provides merely that "an undertaking which prepares" such statements should be deemed a parent undertaking. An undertaking that previously prepared consolidated financial statements in accordance with the Accounting Directive, but subsequently prepared at least one unconsolidated financial statement (because, for example, it no longer qualified as a "parent undertaking" under the Accounting Directive as enacted in its Member State), should not be deemed a "parent undertaking" under the Recast EIR. Similarly, any undertaking that is obligated under a Member State's national law to prepare consolidated financial statements, but has not yet done so, should not qualify as "parent undertaking" under the Recast EIR.

| A group of companies consisting of "undertakings"
In accordance with its definition, a "group of companies" under the Recast EIR consists of a parent "undertaking" and its subsidiary "undertakings". However, because "undertaking" is an economic concept that does not necessarily prescribe certain types of legal forms in which businesses may be conducted, the use of the term can be problematic. 95 In order to address this problem, the Accounting Directive specifically lists the kinds of "undertakings" per Member State to which it applies (for example, capital companies, such as the Dutch besloten vennootschap and the German GmbH, and specific partnerships). 96 The Recast EIR, by contrast, does not limit its scope to certain types of "undertakings". Instead, the Recast EIR can apply to companies, legal persons and natural persons with or without an independent business or professional activity. 97 In light of the Recast EIR's broader general scope, it would not be advisable to incorporate the Accounting Directive's limitations regarding the types of "undertakings" that may form a "group of companies" when construing the Recast EIR. 98 Himmer has argued that, from a teleological perspective, "undertaking" should be understood to mean any "economically active legal entity", whether it is a natural or legal person. 99 Such a broad reading of "undertaking" would support the objective of efficiently conducting insolvency proceedings: the need for efficiency applies to all types of group insolvencies, regardless of the legal form of the group members. Other interpretations of the meaning of "undertakings" in the Recast EIR support such a broad view: both the German and Dutch translation of "undertaking" ("Unternehmen" and "onderneming") similarly describe an economic concept that disregards the legal form in which the business enterprise is conducted. 100 UNCITRAL's Model Law on Groups maintains a similar broad interpretation. The "Enterprise", the Model Law on Groups' equivalent to the Recast EIR's "undertaking", is defined as "any entity, regardless of its legal form, that is engaged in economic activities and may be governed by the insolvency law". 101 The Recast EIR, however, does provide several limitations as to what may constitute an "undertaking". The first limitation follows from Chapter V's objectives: the "undertakings" that constitute the "group of companies" should all have separate legal personality, 102 as group insolvency law generally aims to deal with the challenges of separate legal entities that form an economically integrated group. 103 An "establishment" on itself, as defined in Article 2(10) of the Recast EIR, will thus not qualify as a separate "undertaking" within the meaning of Articles 2 (13) and (14) of the Recast EIR. Rather, the legal entity as such does. 104 Secondly, the Recast EIR generally excludes certain types of financial undertakings from its scope. 105 Such undertakings can therefore not qualify as "undertakings" within the meaning of Article 2 (13) and (14) of the Recast EIR. Thirdly, the Recast EIR only applies to "insolvency proceedings" as listed in Annex A to the Recast EIR, 106 and therefore, the relevant "undertakings" must be subject to one of the proceedings listed in that Annex.
Accordingly, "undertaking" as used in Articles 2 (13) and (14) of the Recast EIR should be understood to mean any company, legal person, or natural person that conducts an independent business or professional activity, to the extent that they have a legal personality separate from the other relevant undertakings and are subject to one of the "insolvency proceedings" as listed in Annex A to the Recast EIR.

| A broad and flexible approach to defining groups
In the foregoing sections, a broad and flexible approach to the definition for a "group of companies" has been proposed, which would place many different types of groups within the scope of Chapter V. This approach is warranted, because the provisions regarding "groups of companies" in the Recast EIR have very limited impact and are largely directed at consensual communication, cooperation and coordination. 107 In cases where it is evident that there is no benefit in cooperating and coordinating the conduct of insolvency proceedings involving members of a "group of companies", making this determination would entail little time and resources. On the other hand, if communication, cooperation and/or coordination is beneficial to the insolvency proceedings, there would be no harm in obligating insolvency practitioners and courts to conduct themselves in such a manner.
The Recast EIR currently prescribes a "one-size-fits-all" approach to defining groups of companies and in applying the provisions of Chapter V: it makes no distinction on the basis of the degree of integration or centralization. In light of the non-intrusive nature of Chapter V's provisions, this is not problematic. However, if EU insolvency law at a later point in time includes additional regulations governing group insolvencies, it may be advisable to apply a two-tier approach to the provisions regarding groups.
The provisions regarding cooperation and coordination should still apply to a broad array of groups in keeping with the broad and flexible approach to the general definitions of groups discussed above. Provisions that have a more significant impact on the position of stakeholders, such as an involuntary group coordination proceeding, the appointment of a single insolvency practitioner, or the commencement of insolvency proceedings before a single court, should be available only to groups of companies with a certain degree of integration (e.g., business or asset integration), taking into consideration the relevant factors for assessing the level of integration. 108 Moreover, depending on the relevant instrument under which a group member is organized, it may be sensible to assess whether certain provisions should apply only to groups that were centrally managed prior to the commencement of insolvency proceedings. Such limitations could be included in the relevant provisions themselves.
3.2 | Further limitations to the scope of Chapter V 3.2.1 | Multiple proceedings regarding multiple debtors in multiple member states The Recitals to the Recast EIR provide further material and territorial limitations to Chapter V's scope, with Recital 62 stipulating that 109 "The rules on cooperation, communication and coordination in the framework of the insolvency of members of a group of companies provided for in this Regulation should only apply to the extent that proceedings relating to different members of the same group of companies have been opened in more than one Member State". (author's emphasis) As an initial observation, application of Chapter V requires that the relevant insolvency proceedings relate to different members of the group of companies. If a group consists of two companies and a main and a secondary insolvency proceeding are commenced for just one of the two companies pursuant to Article 3 of the Recast EIR, Chapter V does not apply, because the proceedings relate to the same member of the group. 110 Secondly, Recital 62 also establishes a territorial limitation: the provisions included in Chapter V apply in a cross-border case only where the proceedings regarding different group members are opened in more than one Member Stateexcluding Denmark. 111 The mere existence of any cross-border element (such as a group member having assets in another Member State) is insufficient. As a result, if a group, for example, consists of two companies established in different Member States, but insolvency proceedings regarding these two companies are only opened in one Member State, 112 the provisions in Chapter V do not apply. 113 In cases where all group members' insolvency proceedings have been commenced in the same Member State, it is up to the relevant Member State's national insolvency law to provide rules on the conduct of those insolvency proceedings, if any. 114 The Recast EIR explicitly allows Member States to establish national rules regarding cooperation, communication and coordination in connection with insolvencies of group members to supplement Chapter V, provided that the scope of those national rules is limited to the particular Member State's jurisdiction and that their application would not impair the efficiency of the rules set forth in the regulation. 115 If both a Member State's rules and Chapter V apply to (certain) group members, the Member State's rules would be limited in their application in lieu of Chapter V's provisions. 116 If, however, insolvency proceedings involving more than one group member have been opened in more than one Member State, the provisions of Chapter V should also apply with respect to insolvency proceedings regarding group members that are located in the same Member State. Neither Recital 62 nor any other provision of the Recast EIR indicates that Chapter V would not govern proceedings involving group members that are located in the same Member State, provided all other requirements for the application of the Recast EIR have been satisfied. Any other interpretation would arguably undermine the effectiveness of Chapter V.

| Groups with non-member-state undertakings
For a group to qualify as a "group of companies", it is not required that the parent undertaking and all other group members have their COMIs within the EU. 117 However, this does not mean that the provisions of Chapter V apply to all members of that group. Like the Original EIR, the Recast EIR applies only to insolvency proceedings involving debtors whose COMI is located in a Member State. 118 Group companies with a COMI located in a non-Member State thus cannot be subjected to insolvency proceedings under the Recast EIR.
If the COMI of one or more group members is located in a non-Member State, it is unclear to what extent the provisions of Chapter V govern the relationship between the insolvency proceedings opened in a Member State pursuant to the Recast EIR and proceedings commenced in a non-Member State. It has long been argued that the Original EIR did not in any way regulate the cross-border effects of insolvency proceedings in relation to non-Member States, 119 a statement that has also been repeated regarding the Recast EIR. 120 This argument originated from the Explanatory Report to the 1995 Convention on Insolvency Proceedings (the Virgós-Schmit Report), 121 which explicitly provided that the Insolvency Convention only dealt with the "intra-Community effects" of insolvency proceedings in the relations between the Member States and did not regulate the effects of insolvency proceedings vis-à-vis third States. 122 It would be up to the individual Member States to provide appropriate conflict rules.
Based on the CJEU's 2014 Schmid v. Hertel ruling, 123 however, it can no longer be argued that the Recast EIR and its provisions cannot have any effect in relation to (parties that are located in) non-Member States. This case concerned the question whether the court in Germany, the Member State in which the relevant insolvency proceeding was opened, had jurisdiction on the basis of Article 3 of the Original EIR to decide an action to set a transaction aside by virtue of insolvency, given that that action was brought against a person whose place of residence was in Switzerland, a non-Member State. In deciding that the German court had jurisdiction, the CJEU among other things considered that 1. the Original EIR did not contain any general and absolute condition that it only applies to situations involving connecting factors with two or several Member States 124 ; 2. where some provisions expressly prescribe connecting factors with the territory or legal system of at least two Member States (e.g., Article 5(1) of the Original EIR, now Article 8(1) of the Recast EIR), other provisions do not (e.g., Articles 6 and 14 of the Original EIR, now Articles 9 and 17 of the Recast EIR) 125 ; and 3. where the Original EIR's provisions do not expressly prescribe a cross-border element involving at least two Member States, the objectives pursued by the Regulation as resulting in particular from the recitals (e.g., improving the efficiency and effectiveness of insolvency proceedings having cross-border effects, Recital 8 in both the Original and Recast EIR), likewise do not support an interpretation of the Regulation's scope that would be limited to intra-Community effects. 126 In light of the CJEU's general wording in the Schmid v. Hertel ruling and, because the Recast EIR has not been amended in this respect, the Recast EIR's provisions can be deemed to have effects in relation to non-Member States, to the extent that the relevant provisions of the Recast EIR and their objectives do not object against such extra-Community effects.
Himmer has argued, with reference to the Schmid v. Hertel ruling, that the provisions of Chapter V also apply in relation to insolvency proceedings which are opened in non-Member States, both in the sense that EU insolvency practitioners and courts are also allowed to communicate and cooperate with insolvency practitioners and courts in insolvency proceedings originating from a non-Member State, as well as granting rights to such extra-Communal parties. 127 As a consequence, Himmer argues that the EU legislator should either limit the applicability of these provisions to cases of reciprocity or include protective clauses which limit the application of Chapter V's provisions in relation to non-Member States to cases where such application is in the interest of the stakeholders in the European insolvency proceedings. 128 Contrary to what Himmer argues, the provisions of Chapter V expressly exclude insolvency proceedings originating from non-Member States from their scope. Chapter V refers to "insolvency proceedings", "insolvency practitioners", and "courts", which are all defined within the Recast EIR's framework and expressly limited to proceedings opened in Member States, 129 certain practitioners appointed by Member States, 130 and courts of Member States. 131 Therefore, while the Recast EIR in general can have extra-Community effects, the provisions on group insolvencies are limited to intra-Community relations between insolvency proceedings of group members, as a result of these definitions.
Taking the group in the organization chart below as an example, Article 56 of the Recast EIR (assuming it applies) imposes an obligation on the insolvency practitioners of Dutch Holding BV and German GmbH to cooperate with each other. However, the insolvency practitioners are not obligated to cooperate and communicate with the insolvency practitioner of Swiss GmbH, as he or she does not qualify as an "insolvency practitioner". 132 Moreover, as the Recast EIR in general does not apply to the insolvency proceeding of Swiss GmbH, Article 56 of the Recast EIR does not impose any obligation on Swiss GmbH's insolvency practitioner to cooperate with the insolvency practitioners of Dutch Holding BV and German GmbH. Whether the Swiss insolvency practitioner is under an obligation to cooperate with them is a matter of Swiss law. The same reasoning applies for instance, mutatis mutandis, regarding participation in a GCP.
Chapter V's objectives are, however, best served by a broad territorial application of its provisions: if the relevant group members conducted their business in an integrated manner and/or were centrally managed prior to the insolvency proceedings, there will likely be efficiency benefits and surplus value realized if their proceedings are coordinated, regardless of where those insolvency proceedings were opened. As mentioned previously, UNCITRAL has recently developed the Model Law on Groups, which contains provisions similar to Chapter V, and with its "planning proceeding", even contains a meta proceeding similar to the GCP. If the Recast EIR is amended in the future, states may already have enacted the Model Law on Groups as part of their national insolvency laws.
C H A R T 2 Sample organizational chart. Source: Author's own It could therefore be advisable for the EU legislator to revisit the relationship between Chapter V and non-Member States' national laws on cross-border group insolvency proceedings. That could, for example, be done by allowing Member States' courts and insolvency practitioners to also cooperate and coordinate with their counterparts in insolvency proceedings regarding group companies that originate from non-Member States, to the extent that such cooperation and coordination is appropriate to facilitate the effective administration of the Member States' proceedings.

| PECULIARITIES REGARDING HETERARCHICAL GROUPS
The Recast EIR's definitions for groups of companies are formulated according to a typical pyramid structure: an (ultimate) parent company which (directly or indirectly) controls its subsidiaries. This approach excludes certain types of groups for which cooperation and coordination amongst group members in insolvency proceedings could be beneficial. This is exemplified by heterarchical groups such as the German Gleichordnungskonzern. A heterarchical group, as such, does not fall within the definition of "groups of companies" in accordance with the first sentence of Article 2(14) of the Recast EIR, because there is no parent undertaking controlling a subsidiary: the group companies operate on an equal basis. A heterarchical group will nonetheless be considered a "group of companies" under the Recast EIR, if it prepares consolidated financial statements in accordance with the Accounting Directive. This is significantly more likely in Member States that have opted to extend the consolidation requirement to heterarchical groups, such as Germany. Under the laws of Member States that chose not to implement similar provisions pursuant to Article 22(7) of the Accounting Directive, 133 a heterarchical group would not be considered a "group of companies" under the Recast EIR.
Even to the extent that Member States have chosen to implement a consolidation requirement for heterarchical companies, the Accounting Directive exempts a significant portion of "groups" from its application. Small-sized groups are exempted and Member States may elect to also exempt medium-sized groups. 134 As such, groups with as many as 250 employees, a balance sheet total of EUR 20 million and/or a net turnover of EUR 40 million may be relieved from the obligation to prepare consolidated financial statements. 135 In the absence of any obligation to do so, heterarchical groups that fall within these parameters will often not qualify as a "group of companies" under the Recast EIR. This restricted scope of groups will detract from Chapter V's effectiveness.

C H A R T 3 Sample organizational chart.
Source: Author's own If, for instance, a single natural person is the shareholder and director of a German GmbH and a Dutch BV that conducted an integrated business prior to commencing insolvency proceedings (see the organization chart above), but the companies did not prepare consolidated financial statements, the group would not qualify as a "group of companies" under the Recast EIR, because the group functions without a controlling "parent undertaking". 136 As a result, insolvency practitioners and courts in the insolvency proceedings regarding these companies will not be obligated by Chapter V to cooperate and coordinate their conduct and will not benefit from Chapter V's provisions, such as the right to participate in other group members' insolvency proceedings.
If the heterarchical group operated in an integrated manner prior to the insolvency proceedings, this would be contrary to the objectives of efficient management and might result in loss of synergy value. 137 This becomes even more apparent if a second "layer" of group companies is added and the heterarchical group is combined with several subordinated group companies (as shown in the organization chart below). The Recast EIR would treat this group as two "groups of companies", with the German GmbH as the "parent undertaking" of subsidiaries German 2 GmbH and Austrian GmbH, and Dutch BV as the "parent undertaking" of subsidiaries Dutch 2 BV and Belgium BV. The provisions of Chapter V would thus not apply, for example, to the German and Belgian insolvency proceedings, but would apply to the German and Austrian insolvency proceedings.
The German legislature explicitly chose to expand the domestic German definition for Unternehmensgruppe beyond hierarchical subordination groups to include Gleichordnungskonzerne. 138 Therefore, the definition for Unternehmensgruppe includes not only groups of companies that are C H A R T 4 Sample organizational chart. Source: Author's own directly or indirectly connected through the potential for exercising a dominant influence, but also groups of companies that are connected, because they "are managed on a unified basis" (Zusammenfassung unter einheitlicher Leitung) 139 language that also appears in the Seventh Council Directive and the Accounting Directive. 140 Such unified management can, for instance, exist where the group members' management overlaps in whole or in part 141 or when the group members have a common sole or majority shareholder. 142 In light of the above-mentioned types of (partially) heterarchical groups that currently fall outside the scope of Chapter V, if the Recast EIR is amended, it would be advisable to include groups that are managed on a unified basis in the regulation's definition of "groups of companies", similar to the German definition for Unternehmensgruppe. This would allow (partially) heterarchical groups to qualify as a "group of companies", as was intended at the inception of the EU's legislative deliberations concerning the Recast EIR. 143

| CONCLUSION: A GOAL-ORIENTED APPROACH TOWARDS DEFINING GROUPS OF COMPANIES
This article set out to examine the scope of the Recast EIR's provisions on group insolvency included in Chapter V. In light of the specific objectives of the Recast EIRand more specifically, Chapter Vthe definition for a "group of companies" under the Recast EIR should be interpreted independently, rather than in conformity with other EU law instruments such as the Accounting Directive. The broad and flexible interpretation of "group of companies" based upon such a teleological approach would include a large number of groups within the scope of the Recast EIR's provisions regarding group insolvencies. Given the non-intrusive nature of the provisions on group insolvency currently included in the Recast EIR, such an interpretation is warranted.
With its current scope, Chapter V covers many, but not all, common types of groups. In defining a "group of companies", the EU legislature elected to limit the provisions on group insolvency to subordination groups. Because stakeholders of heterarchical groups would also benefit from coordinated conduct in insolvency proceedings, a similarly broad approach towards the presumption included in Article 2 (14), 2nd sentence, of the Recast EIR is proposed. Where such groups currently fall outside the scope of Chapter V, an amendment to definition of a "group of companies" is proposed to also include them.