Table of content
I. Problem outline.
II. Operational context of an umbrella clause.
III. General effect of an umbrella clause.
1. Arguments on the general effect of an umbrella clause.
a) The ‘ordinary meaning’ argument
b) The ‘exception’ argument
c) The ‘object and purpose’ argument
d) The ‘floodgates’ argument
e) The ‘superfluousness’ argument
f) The ‘location’ argument
g) The ‘exclusive dispute settlement clause’ argument
2. Alternative reading based on the ‘equivalence of application’ argument?.
3. Intermediate result on the general effect
IV. Scope of application of an umbrella clause.
1. Ratione Materiae.
a) Obligations covered.
b) Determination of a breach.
2. Ratione Personae.
In recent years umbrella clauses have become a frequent vehicle for safeguarding investment contracts and their interpretation a major challenge for the competent tribunals. In contrast to the common view umbrella clauses do not blur the distinction between international and municipal law provided that exactly this fundamental distinction is observed in the process of interpretation. This paper advocates a restrictive interpretation of umbrella clauses in order to achieve a balance between the interests of involved actors. In line with the wording and the original function of investment treaties it is argued that only sovereign conduct of the host state may constitute a violation of the umbrella clause obligation.
I. Problem outline
An umbrella clause (also known as ‘mirror’, ‘observance of undertakings’, ‘observation’, ‘pacta sunt servanda’, ‘sanctity of contract’ or ‘respect’ clause) is a provision in an international investment agreement (IIA) by which the host state guarantees the observation of obligations assumed by it with respect to investments. The name stems from the common notion that by means of this clause contracts and other obligations are put under the treaty’s protective umbrella. It was not until a decade ago that both general effect and exact scope of umbrella clauses were questioned helping them to the prominence they enjoy nowadays. Notably, however, umbrella clauses of any type are only contained in approximately 40 per cent of all bilateral investment treaties (BIT).
Particular challenge with respect to procedural and academic discourse on umbrella clauses arises from the common presentation – especially when dealt with umbrella clauses en passant – that there would be two approaches to their interpretation: A traditional wide view contending that due to an umbrella clause a breach of contract can be claimed under treaty and therefore reviewed by an international tribunal and a narrow one that does not attribute to it the leverage to elevate a contract claim on treaty level. This generalised picture is at least misrepresenting since it mixes up the questions regarding the general effect of an umbrella clause, its scope of application and how it correlates with jurisdictional issues. Different decisions and inconsistent reasoning of the tribunals - that are not bound by previous findings - not only do not help confusion but are jointly responsible for it.
Starting with the first award dealing directly with the application of an umbrella clause in the case SGS v Pakistan attempts have been made to restrict the effect of such clauses concerned with the fact that a wide interpretation might lead to an improper advantage of the investor at the expense of the host state. However, undesirable effects alone do not make an interpretation ‘wrongful’. This paper seeks to pin the articulated concerns to legal arguments. In order to restore the argumentative order, three analytic stages must be distinguished: (1) the operational context of an umbrella clause; (2) its actual general effect; and (3) the scope of operation of the particular provision.
Significant analytic difficulty is, however, that the exact wording of an umbrella clause might differ substantially from treaty to treaty. The following study is based on a common comprehensive (‘proper’) umbrella clause reading:
“Each Contracting Party shall observe any obligation it may have entered into with regard to investments.”
II. Operational context of an umbrella clause
The context in which an umbrella clause operates is the legal framework of an investment. For the purposes of this paper it is sufficient but essential to distinguish the international law level represented by the IIA and the municipal law level represented by an investor-state contract (‘state contract’) or other obligation. Whereas the former provides for a general legal framework for any eligible investment the latter stipulates specific terms and conditions of a particular investment. Hence, each investment is governed by two independent regimes and a breach may occur on each level. On the other hand, an alleged action may be both a breach of the municipal law obligation and a violation of the treaty and can therefore be prosecuted on both levels resulting in a contract claim and a treaty claim that are separate and coexistent.
Decisively however, under customary international law a breach of a contract by a state does not by itself trigger that state’s responsibility on the international law level. Only if the alleged conduct itself constitutes a breach of a treaty provision it may give rise to a treaty claim . Hence, the effect of the umbrella clause concerns a sub-question to the general question of whether a breach of contract at the same time violates the treaty. Therefore, an umbrella clause by no means transforms, converts or transubstantiates contractual claims into treaty claims or contractual undertakings into international law obligations but can only give rise to an additional and independent claim based on alleged violation of the treaty (i.e. the umbrella clause).
The second important issue is not to mix up the question of legal effect of an umbrella clause with the question of whether a tribunal has jurisdiction over contract claims. Investor-state contracts usually contain a dispute resolution clause referring all disputes with regard to the contract to a particular forum, often the domestic courts. IIAs on the other hand often contain a very broad dispute settlement clause referring all disputes with respect to the investment to a selected forum. Thus, one is easily apt to call upon the umbrella clause to ease the conflict between the two forum provisions in favour of the treaty selection. However, keeping in mind the distinction between contract and treaty claims it is easy to determine that the forum selection clause in the contract can only, if at all, oust the tribunal’s jurisdiction over contractual claims. By contrast, the treaty claim is unaffected. Thus, since an umbrella clause only concerns the treaty claim, it cannot have any effect with respect to the forum of the contract claim.
Summarizing this paragraph the following can be concluded: It is crucial to distinguish between municipal law and treaty claims. An umbrella clause does not have a transformative effect. It ‘only’ constitutes an independent substantive obligation under an IIA . Therefore the so-called ‘elevator’ effect is limited to the question of whether a breach of contract is a breach of the umbrella clause and therefore amounts to a breach of treaty.
III. General effect of an umbrella clause
The so settled operational context of an umbrella clause does not, however, determine the actual effect of that clause. Traditional view has always been that an umbrella clause operates in a way by which a breach of contract amounts to a breach of treaty. However, this position was contested by SGS v Pakistan tribunal which, in turn, was shortly followed by a diametrically opposing decision in SGS v Philippines. Subsequently, both decisions gained support and the arguments worked out by the tribunals continue to shape the dispute.
1. Arguments on the general effect of an umbrella clause
Generally speaking interpretation means objective analysis of a provision. And since an umbrella clause forms part of an IIA its interpretation is governed by Articles 31 and 32 of the Vienna Convention on the Law of Treaties ( VCLT). Only few tribunals have acknowledged this explicitly or by implication.
 Andrew Newcombe and Lluís Paradell, Law and Practice of Investment Treaties – Standards of Treatment (The Netherlands, Wolters Kluwer International, 2009) 443 with footnote 29. An umbrella clause could possibly be invoked in other treaties on the basis of a most-favoured-nation clause, see Impregilo v Pakistan, para 223.
 See particularly Consorzio v Algeria, para 25(ii); Ioana Tudor, The Fair and Equitable Treatment Standard in the International Law of Foreign Investment (New York, Oxford University Press, 2008) 193 et seqq; Jacomijn J van Haersolte-van Hof and Anne K Hoffmann, “The Relationship between International Tribunals and Domestic Courts” in Peter Muchlinski, Federico Ortino and Christoph Schreuer (eds.), The Oxford Handbook of International Investment Law (New York, Oxford University Press, 2008) 962, 979-980; Kaj Hobér, “State Responsibility and Attribution” in Peter Muchlinski, Federico Ortino and Christoph Schreuer (eds.), The Oxford Handbook of International Investment Law (New York, Oxford University Press, 2008) 549, 575.
 El Paso v Argentina, paras 63-88 and Pan American v Argentina, paras 94-116 serve as an excellent example of a complete misunderstanding of the umbrella clause concept.
 Vivendi v Argentina (Annulment), paras 95-96; SGS v Pakistan, para 147; Noble Ventures v Romania, para 53.
 Andrew Newcombe and Lluís Paradell, above No 1, 439; Stephan W. Schill, “Umbrella Clauses as Public Law Concepts in Comparative Perspective” in Stephan W. Schill (ed.), International Investment Law and Comparative Public Law (New York, Oxford University Press, 2010) 317, 323-324. See also the clear distinction in Hamester v Ghana, paras 328-330 and Noble Ventures v Romania, para 53. A short overview of related cases is provided by Stanimir A. Alexandrov, “Breaches of Contract and Breaches of Treaty – The Jurisdiction of Treaty-based Arbitration tribunals to Decide Breach of Contract Claims in SGS v. Pakistan and SGS v. Philippines” (2004) 5 Journal of World Investment and Trade 555, 559-560.
 Jarrod Wong, “Umbrella Clauses in Bilateral Investment Treaties: of Breaches of Contract, Treaty Violations, and the Divide between Developing and Developed Countries in Foreign Investment Disputes” (2006) 14 George Mason Law Review 135, 172. This aspect is misunderstood in El Paso v Argentina, paras 83-87; Pan American v Argentina, paras 111-116 and BIVAC v Paraguay, para 149 even though the claimants explicitly argued this issue in each case.
 This understanding in SGS v Pakistan, para 172; El Paso v Argentina, para 70; Pan American v Argentina, para 99.
 This understanding in SGS v Philippines, para 117; El Paso v Argentina, para 67; Pan American v Argentina, para 96; LG&E v Argentina, para 174. See also Noble Ventures v Romania, para 46, though with a correct picture of the effect in para 85.
 Jarrod Wong, above No 6, 170 and 172. See also the clear distinction in Impregilo v Pakistan, para 258.
 This happens e.g. in Eureko v Poland, para 250; El Paso v Argentina, paras 65-67 and 70.
 Vivendi v Argentina (Annulment), para 101; Eureko v Poland, paras 97-114; CMS v Argentina (Jurisdiction), para 76; Stanimir A. Alexandrov, above No 5, 564.
 Stephan W. Schill, above No 5, 331; Thomas W. Wälde, “The “Umbrella” Clause in Investment Arbitration – A Comment on Original Intentions and Recent Cases” (2005) 6 Journal of World Investment and Trade 183, 213; Christoph Schreuer, “Travelling the BIT Route – Of Waiting Periods, Umbrella Clauses and Forks in the Road” (2004) 5 Journal of World Investment and Trade 231, 250; Noble Ventures v Romania, para 51; implied in Hamester v Ghana, para 324.
 This effect is correctly presented in Noble Ventures v Romania, para 85.
 See Christoph Schreuer, above No 12, 250-251; Stanimir A. Alexandrov, above No 5, 566-569.
 In support of SGS v Pakistan explicitly: Joy Mining v Egypt, para 81; El Paso v Argentina, para 82; Pan American v Argentina, para 110. In support of SGS v Philippines explicitly: Eureko v Poland, para 257.
 J. Romesh Weeramantry, Treaty Interpretation in Investment Arbitration (Oxford, Oxford University Press, 2012) para 6.99; Jarrod Wong, above No 6, 173.
 Whether or not the treaty partners are party to the Convention because it states in a declaratory manner what is customary international law anyway, see J. Romesh Weeramantry, above No 16, para 1.17.
 Eureko v Poland, para 247; Noble Ventures v Romania, para 50; Enron v Argentina, para 273; Siemens v Argentina (Jurisdiction), para 80.
 By paraphrasing Art 31(1) e.g. SGS v Pakistan, para 164.