In December 2015, I published an article examining whether there was a trend towards the elimination of umbrella clauses from investment agreements, be they bilateral, multilateral, or model investment treaties. By that time, model bilateral investment treaties (BITs) from the United States, France, Canada, Colombia, and the Southern African Development Community (SADC) and many prominent multilateral investment agreements, including the NAFTA and the ASEAN-Australia-New Zealand FTA, did not contain umbrella clauses. The position against umbrella clauses was clear.
Furthermore, the much anticipated Trans-Pacific Partnership (TPP) and EU-Canada Comprehensive Economic and Trade Agreement (CETA) showed signs that they would not contain umbrella clauses, as well as other important jurisdictions in updating their model BITs, such as India and Norway.
The landscape today
According to the United Nations Conference on Trade and Development (UNCTAD), 52 new investment agreements have been signed since December 2015. In addition, the model BITs from Norway and India were approved. Thus, we have 54 new instruments relating to investment protection. Of these 54 instruments, only two (3.7%) contain an umbrella clause, namely, the Austria-Kyrgyzstan BIT and the Japan-Iran BIT.
While it is not known whether there are more investment agreements that have not yet become public, the trend seems clear and it can be explained by the comments made in 2007 to the draft Norwegian model BIT: “[t]he point of departure for the work on a new model agreement has been that the Arbitration Tribunal shall only be able to consider alleged breaches of the standards in the interstate investment agreement.” This comment reflects the main concern with umbrella clauses: the difficulty to determine whether or not they can work as a “bridge” to bring claims arising from contractual relations into the sphere of investment treaty protection. This concern is shown in the variety of forms and scenarios in which the application of umbrella clauses is an issue:
- The difference between contract claims and treaty claims;
- The requirement of exercising sovereign authority (puissance publique) which has the effect of breaching a contract;
- Whether the umbrella clause supersedes a forum selection clause contained in a public contract;
- Whether shareholders and parent companies that did not sign the public contract can benefit from the umbrella clause;
- Whether sub-state entities that signed the public contract are covered by an umbrella clause.
The issue of inconsistency
Each of these issues has been addressed by arbitral tribunals, which have provided different answers. Take for example the landmark SGS cases against Pakistan (ARB/01/13), the Philippines (ARB/02/6), and Paraguay (ARB/07/29). Each of these cases dealt with almost identically worded umbrella clauses; however each tribunal interpreted the umbrella clause in a different way. In SGS v. Pakistan, the tribunal held that “in the face of a valid forum selection clause,” there was no need to elevate claims grounded in a contract to treaty claims. Just a year later, the tribunal in SGS v. Philippines held that an umbrella clause “provide[s] assurances to foreign investors with regard to the performance of obligations assumed by the host State under its own law with regard to specific investments”; however, it decided to stay the arbitral proceedings in order to wait for the Philippine courts to decide the amount of money the government owed SGS. Finally, the tribunal in SGS v. Paraguay adopted a different and broader interpretation, holding that the ordinary meaning of the word commitment in the umbrella clause clearly encompassed contractual obligations, and that the clause “provide[d] no basis for excluding contracts from the scope of ‘commitments’ covered in the [umbrella clause].”
What is even more worrisome is that the set of facts in each of the SGS cases was basically the same: the breach of a service contract for pre-shipment inspection. And over the years different tribunals have followed not one of the interpretations given by the SGS cases, but all three of them, as some examples summarized in the following chart:
|SGS v. Pakistan Interpretation||SGS v. Philippines Interpretation||SGS v. Paraguay Interpretation|
|– Toto Construzioni v. Lebanon
– Salini v. Jordan
– El Paso v. Argentina
– Siemens v. Argentina
– Joy Mining v. Egypt
|– BIVAC v. Paraguay
– Bosh v. Ukraine
|– Eureko v. Poland
– Noble Ventures v. Romania
– Burlington v. Ecuador
– Duke Energy v. Ecuador
The other issues cited above also had different interpretations, adding more fuel to the debate of whether an umbrella clause encompasses contractual claims and, if so, under what circumstances. Unfortunately, a consensus has not been reached, and this is showing in treaty practice. While tribunals do struggle in finding the correct interpretation of other treaty standards like fair and equitable treatment, full protection and security, and expropriation, these standards establish clear obligations for states. In contrast, the main purpose of an umbrella clause is more ambiguous: to bring under the “umbrella” of the treaty obligations of the state that arose out of a different instrument, i.e., a contract.
The generic language of umbrella clauses has contributed to intense debates since the first SGS case and continues to do so. Yet, treaty practice shows a trend towards the elimination of the umbrella clause from investment agreements. Notwithstanding this perceived new trend, umbrella clauses are still present in new treaties. In this sense, it is worth mentioning the wording of the one contained in the Austria-Kyrgyzstan BIT, which clarifies that “the breach of a contract between the investor and the host State will amount to a violation of this treaty.” That wording would provide more certainty in adjudicating future cases under this particular BIT; however, it might be too late for others.