Treaty shopping, also called corporate (re-)structuring, is most often associated with legal persons, in particular mailbox companies. Much discussed in this respect is the practice of “round-tripping” where the investor-claimant is foreign-incorporated, but majority-controlled by natural or legal persons of host State nationality (see e.g. Tokios Tokeles v Ukraine, ICSID Case No. ARB/02/18, Decision on Jurisdiction, 29 April 2004). In these cases, arbitral tribunals have virtually uniformly rejected Respondents´ pleas to “pierce the corporate veil” and look for the “real” (upstream) investor in the absence of a corresponding, treaty-based requirement, thus adopting a strictly formal and consent-oriented approach towards the notion of investor. This is so even where the foreign-incorporated investor is controlled by a natural person holding dual nationality, one of which the host State nationality, as was the case in Wena Hotels Ltd v Egypt (ICSID Case No. ARB/98/4, Proceeding on the Jurisdiction, 25 May 1999).
Less discussed, and more difficult in legal terms, is the situation when a locally incorporated investor brings the investment claim and is controlled by a natural person holding dual nationality, one of which the host State nationality. These difficulties stem in particular from the fact that the drafters of the ICSID Convention did not define the notion of “foreign control” in Art. 25 (2)(b)(ii). Thus, it is unclear, and has been inconsistently decided by arbitral jurisprudence, whether “foreign control” should mean purely formal or rather effective control. It is equally unclear whether in a corporate structure stretching over several corporate tiers, the arbitral tribunal should, or even must, enquire into the nationality of the ultimate controller of the investment, even after it has already found a qualifying nationality “further downstream”.
These questions matter in the context of deciding whether a claim brought on grounds of such a constellation would be “legitimate nationality planning” or “using the Convention for purposes for which it was clearly not intended”. Requiring effective as opposed to formal control and/or enquiring into the ultimate controller´s nationality is a more probing test for the claimant and enables tribunals to look more easily for the “true” controller of the investment, thus “unmasking” corporate structures that may have been set up with the aim of treaty shopping (but that, absent fraud, remain perfectly legal constructions).
Two recent decisions suggest that arbitral tribunals are becoming increasingly aware of the treaty shopping-potential lying in this constellation, drawing up new limits.
Burimi SRL and Eagle Games SH.A v Albania
In Burimi SRL and Eagle Games SH.A v Albania (ICSID Case No. ARB/11/18, Award, 29 May 2013), one of the Claimants, Eagle Games, was an Albania-incorporated company whose majority shareholding was held by an Italian-Albanian dual national. Addressing the question of whether this claimant could be considered as being under “foreign control” for the purpose of Article 25 (2)(b)(ii) of the ICSID Convention, the tribunal found that:
[W]hile neither the ICSID Convention nor relevant precedents address the potential for a dual national invoking one of his two nationalities to establish jurisdiction over a claim brought in the name of a juridical person under the second clause of Article 25 (2)(b), it strikes the Tribunal as anomalous that the principle against use of dual nationality in 25 (2)(a) would not transfer to the potential use of dual nationality in 25 (2)(b). Otherwise, any dual national who is a national of the Contracting State to a dispute could circumvent the bar on claims in Article 25 (2)(a) by establishing a company in that state and asserting foreign control of that company by virtue of his second (foreign) nationality.
The tribunal thus dismissed jurisdiction ratione personae over the Albania-incorporated, dual national-controlled claimant.
National Gas Company S.A.E. v Egypt
In the more recent National Gas Company S.A.E. v Egypt case (ICSID Case No. ARB/11/7, Award, 3 April 2014), the claimant was a locally incorporated company controlled by two layers of shell companies incorporated in the United Arab Emirates and ultimately (majority-)controlled by an Egyptian-Canadian dual national. The arbitration being brought under the ICSID Convention, Egypt argued that the claimant was not under foreign control within the meaning of Article 25 (2)(b)(ii) given the (dual) host State nationality of the ultimate controller.
Reviewing relevant prior jurisprudence, the tribunal confirmed that Article 25 (2)(b) (ii) established both a subjective and an objective test that had to be met cumulatively for jurisdiction to exist. The subjective test concerned the agreement of the parties to treat a locally incorporated company as a national of the other Contracting Party. This requirement was fulfilled in the case at hand as Article 10 (4) of the Egypt-United Arab Emirates BIT contained this agreement. However, the agreement on treating a local company as foreign not being sufficient on its own, the question was whether there was foreign control objectively speaking. In this regard, the tribunal recalled that while such an agreement could create a “rebuttable presumption” as to the existence of foreign control, limits existed if the ICSID Convention was to be used “for purposes for which it was clearly not intended”.
Applying an object and purpose-based interpretation of the ICSID Convention, the tribunal found a:
… significant difference under Article 25(2)(b) between (i) control exercised by a national of the Contracting State against which the Claimant asserts its claim and (ii) control by a national of another Contracting State. The latter situation violates no principle of international law and is consistent with the text of the ICSID Convention. On the other hand, the former situation violates the general limitation in Article 25(1) and the first part of Article 25(2)(b) of the ICSID Convention in regard to both Contracting States and nationals (including dual nationals). In other words, the latter is consistent with the object and purpose of the ICSID Convention; but the former … would permit the use of the ICSID Convention for a purpose for which it was clearly not intended and it would breach its outer limits … Article 25(2)(b) operates only as a qualified exception to the general limitation to ICSID jurisdiction in Article 25: a sardine cannot swallow a whale.
Endorsing the reasoning in the Burimi case, the National Gas tribunal considered the above to constitute “established principles with a clear dividing-line, making for legal predictability and certainty”, dismissing the case without for that matter considering the claim as abusive.
The Burimi and National Gas decisions seem to consolidate a line of jurisprudence that increasingly tends to dismiss jurisdiction on the grounds of Article 25 (2)(b)(ii) when the ultimate controller of a local investor-claimant is a host State national (whether single or dual). In the case of (single) host State nationality, this has so far remained controversial because of the lack of any explicit limitation in the Convention, making a dismissal often dependent on the availability of information on ultimate control structures and arbitral discretion. This is different in the case of (dual) host State nationals due to the existence of an explicit “host State bias” in Article 25 (2)(a) of the ICSID Convention. Indeed, the “host State bias” argument arguably might translate into a broader “principle” (at least in ICSID arbitrations) helping to make a more solid legal argument why even in the case of a single host State national exercising ultimate control, jurisdiction ratione personae should be rejected.
More fundamentally, the treatment of claims by locally incorporated entities ultimately controlled by host State nationals shows an important discrepancy compared to the treatment of claims by foreign-incorporated entities ultimately controlled by host State nationals. As is well-known, the interposition of corporate vehicles in the chain of ownership has the potential to enhance treaty protection, in other words to further treaty shopping. The discrepancy becomes most obvious in the case of dual nationals (with host State nationality) ultimately controlling foreign and/or locally incorporated entities. Thus, while a claim by a foreign-incorporated entity controlled by a dual (host State) national does not encounter any problems, as seen in Wena Hotels v Egypt, the claim by a locally incorporated entity ultimately controlled by a dual (host State) national will most likely lead to a dismissal of the case (at least in ICSID arbitrations). In such a chain of ownership, the chances of success for an investment claim to cross (at least) the first (jurisdictional) hurdle therefore largely depend on whether it is the foreign or the locally incorporated entity bringing the claim.
Of course, the drafters of the ICSID Convention had good reasons for treating foreign and locally incorporated investors differently; the legal conclusions, by arbitral tribunals, concerning the claims brought by foreign-incorporated entities can thus not be objected to. Yet, the potential for manipulation lying in this inconsistency is obvious. As a consequence, the use and interposition of mailbox companies in a corporate structure becomes even more attractive, exacerbating potential policy concerns with respect to treaty shopping.