On 14 November 2019, AG Kokott released her opinion in Dong Yang Electronics (Case C-547/18), referred to the Court of Justice of European Union (CJEU) by the Regional Administrative Court of Wroclaw (Poland) and concerning the existence of a VAT fixed establishment (FE) of a third-country (Korea) company in an EU Member State (Poland).
As AG Kokott recalled while introducing the case, the CJEU has previously been asked to determine whether a subsidiary can also be regarded as a FE of a parent company established in another country for EU VAT purposes. However, a clear statement on this matter cannot be found in the existing CJEU’s case law. In Dong Yang Electronics, the Court has thus the chance to get things clear once and for all.
Facts and Questions
The background to the request for a preliminary ruling is that a company established in Korea (LG Korea) commissioned a third-party Polish company (Dong Yang Electronics) to supply assembly services relating to certain materials for processing printed circuit boards provided to it by a Polish subsidiary of the Korean parent company (LG Poland Production). The processing and assembly of the products was undertaken jointly by Dong Yang Electronics and LG Poland Production based on a cooperation agreement, with each undertaking taking care of a specific part of the overall supply to LG Korea. The Polish subsidiary was also tasked with storage and logistics for the finished products owned by the Korean parent company, which then sold those products to another Polish subsidiary (LG Poland Sales), which finally marketed them across the EU. The matter under dispute was whether Dong Yang Electronics – the supplier – provided assembly services to the Korean parent company or to the Polish subsidiary as a FE of the parent company, that is who was the actual beneficiary of the services supplied.
The Polish tax authorities argued that the latter situation applied on the ground that the Korean parent company had itself created a FE in Poland by “exploiting the economy potential” of its subsidiary “through implementing a suitable business model by way of the agreements” concluded with all the parties involved in the overall supply. Based on those findings, the place of supply would be in Poland and Polish VAT would have been incurred pursuant to Article 44 of the VAT Directive (VD). This view was however rejected by the applicant, which contended that the requirements for the existence of FE were not met in the case. This, in short, is the gist of the question submitted to the CJEU.
AG Kokott’s Findings
AG Kokott did not shy away from providing a straightforward answer to the question referred. In her opinion, the mere fact that a parent company established in a third country has a subsidiary in an EU country does not result in the parent company having a FE in the latter country. This conclusion stems directly from the simple evidence that a parent company and a subsidiary are not one taxable person but indeed two legally autonomous entities, although one economically dependent from the other.
More interesting was however to establish whether and, eventually, under which circumstances the human and technical resources (“the infrastructure”) of a taxable person – the subsidiary – can also constitute a FE of a different taxable person – the parent company. The AG adopted a very narrow view on when this situation may indeed occur, mainly based on four considerations:
- the subsidiary has its own legal personality and is therefore also a taxable person in its own right;
- under EU VAT, separate legal personality can be disregarded and a single taxable person can be formed only pursuant to Article 11 VD concerning VAT grouping;
- treating the subsidiary as a FE alters the reverse-charging of VAT to the service recipient pursuant to Article 196 VD, thus creating a VAT liability for a foreign company in the FE country;
- an ex-post recharacterization of a subsidiary as a FE of a foreign company undermines legal certainty, in so far as the other party to a transaction must know whether the subsidiary or the parent company is in fact liable for VAT.
The AG however found a notable exception to the above conclusions, which applies when, in light of the doctrine established by the CJEU in Halifax (Case C-255/02), the chosen contractual relationships amount to an abusive practice. An abusive practice can in particular be found when, through contractual arrangements, a company manages to reduce its own VAT burden, whereas adherence to economic reality would lead to a more burdensome VAT result.
This consideration naturally led the AG to examine the CJEU’s decision in DFDS (Case C-260/95), where the Court found that a subsidiary can actually constitute a FE of the parent company in so far as it acts as a “mere auxiliary organ” of the parent company. The AG however contended that the CJEU’s findings in DFDS were not applicable in the case at hand. This stems from various – although, in the author’s opinion, not all similarly compelling – reasons:
- DFDS concerned a special VAT scheme such as that for tour operators (Articles 306-310 VD), which by itself makes that case’s findings not generalizable;
- in DFDS the subsidiary was, due to the agency agreement in place with the parent-principal, bound to act as an intermediary for the parent company, with the effect that the former could well be equated to a mere auxiliary organ of the latter;
- DFDS concerned a subsidiary as a supplying FE rather than a receiving FE, which means that reference to an auxiliary organ acting for the parent company is not applicable in the latter event;
- DFDS’s findings might apply only when a risk of abuse is concerned, that is if and in so far as the dominant party is able to reduce its own VAT liability through legal arrangements not reflecting any economic reality;
- DFDS has largely been overruled by later CJEU’s decisions, in particular Daimler and Widex (Joined Cases C-318/11 and C-319/11).
Subsidiary and Branch: Same Animals for CIT and VAT?
From an economic perspective, there is no difference between parent-subsidiary structures and head office-branch configurations. Corporate income tax (CIT) mostly reflects this understanding: the head office and the branch are in fact considered as self-standing enterprises to which profits are separately attributed in as much as it occurs in a parent-subsidiary structure. Further, pursuant to Article 7(2) of the OECD Model, fictious transactions – dealings – between the head office and a branch are considered as full-fledged transactions. Also, on a neutrality point of view – a principle at the core of EU VAT – there must be no clearance between the two arrangements for the purpose of such tax. The same idea of neutrality is followed closely in the OECD VAT/GST Guidelines.
For EU VAT purposes, however, such difference in structure does matter. Although VAT in principle applies the same way to both a subsidiary and a FE (e.g. with regard to the right to deduct input VAT), the use of one structure over the other is relevant in so far as a foreign company’s liability to pay VAT in a country depends – inter alia – on whether it has a FE in that country or not. Further, it ought to be recalled that in FCE Bank (Case C-210/04) the CJEU ruled that supplies between a head office and a branch shall be placed outside the scope of VAT. It is all the way different where only the head office or the branch is part of a group of companies established in a country, so that a transaction is deemed to take place between the two for VAT purposes, as it was found in Skandia America (USA) (Case C-7/13).
Towards a “DFDS Exception”?
As rightly pointed out by AG Léger in his opinion in FCE Bank (Case C-210/04) (at para. 64), while the OECD Model applies a legal fiction to head office-branch structures, a fundamental criterion of the EU VAT system is that it is the actual economic situation that matters (for this principle, see e.g. Newey (Case C-653/11), at para. 42). This not only impedes to treat a transaction between the head office and the branch as it had taken place between two independent entities, but also – the author submits, in line with AG Kokott’s reasonings – to treat a subsidiary as a FE of a parent company, unless the actual economic situation suggests otherwise.
More specifically, only when the subsidiary acts as a mere auxiliary organ of the parent company its autonomous legal personality shall be disregarded and the subsidiary is to be treated as a FE of the parent company. For this assessment, the author submits that the parent company must be found as having immediate and constant access to the human and technical resources of the subsidiary (along the same lines, see AG Kokott’s opinion in Welmory (Case C-605/12), at para. 52), in as much as, pursuant to Article 7(7) of the OECD Model, a parent company can have a PE in another country because a space or premises belonging to the subsidiary are at its disposal, even if owned by the latter company. Recalling the “Marks & Spencer exception” concerning definitive losses, we may tentatively refer to this latter scenario as the “DFDS exception”.
It will then be very interesting to see whether the CJEU will follow the conclusions of AG Kokott and thus a “DFDS exception” will indeed be created. Stay tuned.
 In this regard, see Nigl and Others (Case C-340/15), paras. 30-34, where the CJEU maintained that cooperation, even close cooperation, between distinct undertakings does not necessarily result in one person being subordinated to the other.
 Along the same lines, see N. Jovanovic and M.M.W.D. Merkx, Welmory: A Recipe for VAT Avoidance?, 24 EC Tax Review 4 (2015), at p. 208.
 The author notes that the conclusion of AG Kokott in this regard are in line with the sharp criticism expressed by the VAT literature after the CJEU’s decision in DFDS, which points to the situation at stake in that case as exceptional. See, almost unanimously, B.J.M. Terra, Internet and the Concept of “Fixed Establishment” of the Recipient of a Supply of Services: Case C-605/12 (Welmory), 3 World Journal of VAT/GST Law 3 (2014), at p. 217; M.E. van Hilten and H.W.M. van Kesteren, Omzetbelasting (Kluwer 2010), at pp. 85-86, cited by G.-J. van Norden, The Allocation of Taxing Rights to Fixed Establishments in European VAT Legislation, in VAT in an EU and International Perspective (H.V. Arendonk, S. Jansen and R. Paardt eds, IBFD 2011); K. Spies, Permanent Establishment versus Fixed Establishment: The Same or Different?, 71 Bulletin for International Taxation 12 (2017), at p. 714.
 A further confirmation in this regard was found, firstly, in the capital of subsidiary entirely belonging to the parent company and, secondly, in the legal impediment for the subsidiary to provide services to other parties than its parent company.
 The author finds this argument quite unconvincing as long as it claims to distinguish between an “active” and a “passive” FE, an idea which is highly disputed in the VAT literature (see e.g. VAT Committee, Working Paper No. 791. Clarification of the Concept of Fixed Establishment (taxud.c.1(2014)88957)). Against this possibility, following a contextual interpretation method, the author further notes that, pursuant to Articles 14(2)(c) and 28 VD, a taxable person can act in his own name but on behalf of another person either as the supplying or receiving party.
 Notably, AG Kokott argued that in DFDS a risk of abuse was present because the relevant rule in that case provided for taxation where the supplier rather than the customer was established (i.e. based on the origin rather than the destination principle) and, further, due to the fact that that decision concerned “blunt” services rather than services that find their way into physical goods as in the case at hand, which can be more easily monitored. As a matter of fact, in DFDS not only VAT revenues were at stake, but also a risk to distort freedom of competition materialized.
 See, in particular, Guidelines 2.3., stipulating that “VAT rules should be framed in such a way that they are not the primary influence on business decisions”.
 VAT liability is not the only element to consider. For instance, whether or not a foreign company has a FE in a country plays a role in the context of the call-off stock simplification under the new Article 17a VD, since for that provision to apply a foreign supplier can be VAT-registered but not also established in a country.
 See A. Tsielepis, The Devil Is in the Detail: An Analysis of the ECJ’s Attributes to the Fixed Establishment Concept, 28 International VAT Monitor 2 (2017), at p. 213, who, based on the CJEU’s case law (in particular, Welmory), observes that “the prerequisites of an FE of having its own human and technical resources thus seems to be withering”. On a similar strain, see AG Poiares Maduro’s opinion in RAL (Channel Islands) Ltd and Others (Case C-452/03), paras. 47-55.
 Against this background, it ought to be noted that reference to the OECD Model is, by itself, irrelevant, since it concerns direct taxation whereas VAT is an indirect tax (see FCE Bank (Case C-210/04), at para. 39). Nevertheless, the Italian Supreme Court (ISC) still maintains that the concept of FE can be interpreted in light of the PE definition provided for direct taxation in the OECD Model (see ISC, 15 November 2017, n. 27070).