The Concept of “Investment”: Treaty Definitions and Arbitration Interpretations

Abstract

The concept of “investment” as a prerequisite for investor-State arbitration has caused many difficulties and controversies in ISDS practice. In particular, the silence on the term “investment” under the ICSID Convention has given rise to divided interpretations in arbitral decisions on whether the tribunal should appreciate and apply the “ordinary, objective or inherent” meaning of the term “investment” to establish the ICSID jurisdiction. In response to such controversy and uncertainty, many States have used various approaches to narrow and clarify the scope of investment for the avoidance of exposure to ISDS claims, such as the inclusion of the characteristics of investment, the explicit exclusion of certain assets, and the requirement of in accordance with host State’s laws. Nonetheless, the effect of such newer investment treaties remains to be tested in arbitral cases. It is submitted that a “double door” theory may be useful to address the relationship between the Article 25 of the ICSID Convention and the definition of “investment” clause under applicable investment treaties. For the ICSID jurisdiction, the “objective and inherent meaning” of the term “investment” should be appreciated and applied in accordance with Article 31 (1) of the VCLT, which include only the minimum core of the concept. The same ordinary meaning might also be applied in non-ICSID arbitrations only if the governing substantive treaty does not clearly define what constitutes an “investment”. In this regard, the definition of investment clause under investment treaties is likely to subject to greater scrutiny when tribunals apply the “ordinary meaning” of investment in case the substantive investment treaty failed to provide a “special meaning” under Article 31(4) of the VCLT.

Keywords

Introduction

The concept of “investment” is of great importance in both investment treaties and arbitral practice. It determines not only the subject matter coverage of investment treaties that enjoy substantive and procedural rights enshrined in international agreements, but also the jurisdiction ratione materiae for arbitration under Article 25 of the Convention Establishing the International Centre for the Settlement of Investment Disputes (ICSID Convention).

In practice, however, the interpretation of the concept of investment has not been an easy task but subject to intense debates. This is because the ICSID Convention does not define the concept, whilst most investment treaties adopt a broad definition of investment, literally including every kind of assets. As a result, arbitral tribunals have been widely divided particularly on whether the concept of investment in the ICSID Convention and investment treaties should appreciate certain objective criteria or requirements inherent in the term.

Amidst the ongoing reforms of international investment regime, many newer treaties have taken different approaches to minimise State’s exposure to investment arbitration through narrowing the scope of investor-State dispute settlement (ISDS) subject matter and clarifying the definition of investment. Such newer treaty formulations, however, remain to be tested, as the divergence in interpretation of investment carries on in recent arbitral cases.

It is against this backdrop that this chapter evaluates the concept of investment in both investment treaties and investor-State arbitration. Section II reviews the treaty practice on the definition of “investment” highlighting the salient features found in some newer investment treaties. Section III critically evaluates the divergent approaches adopted by arbitral tribunals in the interpretation of the notion of investment in both ICSID and non-ICSID arbitrations. Section IV concludes the chapter.

The Definition of “Investment” in Treaty Practice

Diversity in Investment Definition Clauses

In general, most (if not all) investment treaties provide a definition of “investment”, but the diversity exists in both the types of definition and specific language used in different treaties.

There are two main types of definition of “investment” in the over 3000 investment treaties. Most investment treaties adopt an “asset-based” approach to define “investment” as “every kind of assets”, which is often followed with an illustrative list of protected assets. ¹ Since the categories of assets are provided as examples, any unlisted assets or interests may also be considered as covered investment. The coverage of investment under a traditional asset-based definition can be very broad. Such an open-ended approach to define investment may aim at attracting and promoting inward foreign investment, but it also can expand the covered investment to uncontemplated assets and expose States to investment arbitration beyond expectations.²

By contrast, some investment treaties use an “enterprise-based” approach to define “investment” as assets and interests related to an “enterprise”.³ Despite of the narrower coverage, the enterprise-based definition of investment implies to protect the affiliation or subsidiary of an enterprise established in accordance with the host State’s laws as an independent investment. Additionally, under such a definition, foreign investor can bring claims not only on his own behalf but also on behalf of the enterprise.⁴

Salient Features in Newer Investment Treaties

With the rise of ISDS cases, an increasing number of States are aware of the importance of clarifying the scope of covered investment for minimizing the exposure to investment arbitration and preserving regulatory space. Consequently, States have used different techniques, as detailed below, to refine the definition of investment in newer investment treaties.

Characteristics of Investment

Requiring investment to fulfill specific characteristics has become a prominent feature in many newer investment treaties and a common approach for States to clarify the concept of covered investment. For example, the Switzerland-Egypt BIT (2010) provides that “[t]he term ‘investment’ shall include every kind of assets that has the characteristics of an investment, such as the commitment of capital or other resources, the expectation of gain or profit, or the assumption of risk”.⁵ Similarly, the Austria-Kazakhstan BIT (2010) provide that “…[i]nvestment are understood to have specific characteristics such as the commitment of capital of other resources, or the expectation of gain or profit, or the assumption of risk”.⁶ The Singapore-Myanmar BIT emphasizes, though in a footnote, that “[w]here an asset lacks the characteristics of an investment, that asset is not an investment regardless of the form it may take”.⁷

The specific characteristics of qualified investment may vary in different treaties. The Colombia-UK BIT (2010), for instance, provides that “to qualify as an investment under this Agreement, an asset must have the minimum characteristics as investment, which are the commitment of capital or other resources and the assumption of risk”.⁸ The China-Japan-Korea investment agreement (2012) provides the characteristics of an investment to include “the commitment of capital of other resources, the expectation of gain or profit, or the assumption of risk”.⁹ Similar provisions are found in the Comprehensive and Progressive Agreement for Trans-Pacific Partnership Agreement (CPTPP) and the new United States-Mexico-Canada Agreement (USMCA).¹⁰

Notably, some recent treaties expressly include a certain “duration” as a characteristic of investment. The Morocco-Japan BIT, for instance, provides investment as every kind of asset that “has the characteristics of an investment, such as the commitment or capital or other resources, the expectation of gain or profit, the assumption of risk or certain duration…”¹¹ Likewise, the Morocco-Nigeria BIT provides that the characteristics of investment involves “a commitment of capital or other similar resources, pending profit, risk-taking and certain duration”.¹² In the Iran-Slovakia BIT, the characteristics of investment include “a reasonable duration”.¹³ By contract, the Brazil-India BIT includes “the objective of establishing a lasting interest” in the characteristics of investment, rather than referring to the duration.¹⁴

Under the EU-Canada Comprehensive Economic and Trade Agreement (CETA) , the characteristics of an covered investment include “a certain duration and other characteristics such as the commitment of capital or other resources, the expectation of gain or profit, or the assumption of risk”.¹⁵ In March 2020, the European Union (EU) proposed a draft on the modernization of the Energy Charter Treaty (ECT) where it provided that an asset must have “the characteristics of an investment, including such characteristics as a certain duration, the commitment of capital of other resources, the assumption of risk, or the expectation of gain and profit”.¹⁶ One month later, however, the EU released a revised draft proposal following comments by the EU Member States, in which the EU clarified that “the criteria describing the characteristics of an investment should not all be presented in a facultative manner – as ‘a certain duration’ is indeed mandatory whereas other elements are not and therefore the ‘or’ needs to be replaced by ‘and’”. The changed provision is in line with the CETA text and the EU-Viet Nam Investment Protection Agreement.¹⁷

Another noteworthy development regarding the characteristics of investment is the requirement for an investment to contribute to the (sustainable) development of the host State. For example, the Egypt-Mauritius BIT and the 2015 Indian Model BIT expressly provide the contribution to (sustainable) development as characteristics of investment.¹⁸ Likewise, the India-Kyrgyz BIT provides that the characteristics of an investment include “a significance for the development of the Party in whose territory the investment is made”.¹⁹ The Iran-Slovakia BIT requires the investment to make “an effective contribute to the Host State’s economy”.²⁰

A few recent treaties also refer to the contribution to sustainable development in the definition of investment, though not specify it as a characteristic of investment. For example, the Morocco-Nigeria BIT provides that investment means “an enterprise… together with the asset of the enterprise which contribute sustainable development of that Party and has the characteristics of an investment…”²¹ Notably, the 2019 Morocco Model BIT not only provides that investment means “an asset that over a certain duration, contributes to the sustainable development of the host party”, but also proposes non-exhaustive indictors for measuring the contribution to sustainable development, including: increased production capacity, economic growth, quality of jobs created, duration of the investment, technology transfer, and reduction of poverty.²²

Explicit Exclusion of Certain Investment

Another prominent development and common approach in newer investment treaties is to expressly exclude certain assets from the definition of investment for greater certainty and to avoid extending treaty protections to uncontemplated investments. For example, the USMCA provides that “investment means every asset… that has the characteristics of an investment… but investment does not mean: (i) an order or judgement entered in a judicial or administrative action; (j) claims to money that arise solely from: (i) commercial contracts for the sale of goods or services by a natural person or enterprise in the territory of a Party to an enterprise in the territory of another Party, or (ii) the extension of credit in connection with a commercial contract referred to in subparagraph (j)(i)”.²³ The Canada-Kuwait BIT contains similar exclusions.²⁴

The CETA provides that forms of investment include “claims to money” excluding: (a) claims to money that arise solely from commercial contracts for the sale of goods or services by a natural person or enterprise in the territory of a Party to a natural person or enterprise in the territory of the other Party. (b) the domestic financing of such contracts; or (c) any order, judgment, or arbitral award related to sub-subparagraph (a) or (b)” from the definition of investment.²⁵ The EU-Singapore BIT specifies that “an order or judgement entered in a judicial or administrative action shall not constitute in itself an investment”.²⁶ The CPTPP also expressly excludes order and judgement from the covered investment.²⁷

In a draft proposal for the modernisation of the ECT, the EU specified “claims to money” in line with the CETA text and also proposed to exclude “an order or judgement entered in a judicial or administrative action or an arbitral award” from the definition of investment.²⁸ In a revised proposal, the EU further added clarifying that “a simple loan or financial contribution does not constitute an investment” and “commercial creditors, including lenders or any other third parties to the investment are not protected under the ECT”.²⁹

A few recent treaties have further narrowed the coverage of investment with a more extensive “negative list”. For example, the Kyrgyz-India BIT provides that “investment does not include the following assets of an enterprise: (i) portfolio investments of the enterprise or in another enterprise; (ii) debt securities issued by a government or government-owned or controlled enterprise, or loans to a government or government-owned or controlled enterprise; (iii) any pre-operational expenditure relating to admission, establishment, acquisition or expansion of the enterprise incurred before the commencement of substantial business operations of the enterprise in the territory of the Party where the investment is made; (iv) claims to money that arise solely from commercial contracts for the sale of goods or services by a national or enterprise in the territory of a Party to an enterprise in the territory of another Party; (v) goodwill, brand value, market share or similar intangible rights; (vi) claims to money that arise solely from the extension of credit in connection with any commercial transaction; (vii) an order or judgment sought or entered in any judicial, administrative or arbitral proceeding; (viii) any other claims to money that do not involve the kind of interests or operations set out in the definition of investment in this Treaty.”³⁰ The Brazil-India BIT contains a similar “negative list” to further qualify the coverage of investment.³¹

Likewise, for the avoidance of doubt, the Iran-Slovakia BIT expressly provides that the term “investment” shall not include:

• a) goodwill or market share;

• b) portfolio investment, which is 10% or less shareholding;

• c) claims to money deriving solely from commercial contracts for the sale of goods or services to or from the territory of a Contracting Party to the territory of another country, or to a State enterprise;

• d) futures, swaps, forwards, options, and other derivatives;

• e) assets used for non-business purposes, other than assets of research and development of non-profit organizations;

• f) funds;

• g) the following loans and debt securities:

  • i. debt securities and loans with the original maturity of less than 3 years;

  • ii. a loan to or debt security issued by a financial institution, which is not treated as regulatory capital by the Contracting Party in whose territory the financial institution is located;

  • iii. The extension of credit in connection with a commercial transaction, such as trade financing.³²

Notably, some treaties exclude certain assets from the definition of “investment” on the ground of the lack of characteristics of investment. The USMCA, for instance, provides that “[s]ome forms of debt, such as bonds, debentures, and long-term notes or load, are more likely to have the characteristics of an investment, while other forms of debt, such as claims to payment that are immediately due, are less likely to have these characteristics”.³³ The Japan-Colombia BIT provides a note for the characteristics of investment, specifying that “[e]ach Contracting Party recognizes that some claims to money that (i) are immediately due and result solely from export and import contracts for the sale of goods or services other than such contracts based on orders habitually secured; or (ii) resulted from credit granted in relation with the contracts referred to in subparagraph (i), maturity date of which is less than twelve (12) months; do not have the characteristics of an investment”.³⁴

In Accordance with Local Laws

Although the legality requirement is not a new feature in the definition of investment,³⁵ many recent investment treaties tend to emphasis this requirement as an important method to balance the obligations between host States and investors. For example, the Argentina-Qatar BIT stipulates that the term “investment” means “any kind of asset invested by an investor of one Contracting Party in the territory of the other Contracting Party in accordance with the laws and regulations of the latter Contracting Party, which involves commitment of resources into the territory of the host Contracting Party”.³⁶

The Slovak-Iran BIT provides that “the investment is made and maintained in accordance with the laws of the Host State and in good faith”.³⁷ Similarly, the Kyrgyz-India BIT provides that investments means “an enterprise constituted, organized and operated in good faith by an investor in accordance with the law of the Party in whose territory the investment is made…”.³⁸

In the modernization of ECT, the EU also calls for the legality requirement, though the CETA does not provide for such a requirement in the definition of investment. As suggested by several Member States, the EU proposed that “Investment” shall “includes all investments made in accordance with the applicable law and the domestic law of the host Contracting Party…”³⁹ In particular, the EU emphasized that this requirement is “indeed necessary to ensure that there is a requirement that the investment is also legal under domestic law at the time of (especially as there is little international law applicable at the time of establishment). The inclusion of a reference to domestic law however requires that “CETA language” is added, which clarifies that domestic law can only be taken into consideration as a matter of fact”.⁴⁰

Clearly, both developing and developed countries have taken various approaches to clarify and limit the definition of “investment” in newer investment treaties. The above-mentioned developments are prominent in recent treaty practice, and the measures taken by developing States tend to be more drastic and restrictive than developed States. Such changes are prompted by the increasing ISDS cases, for the avoidance of doubts on the interpretation of “open-ended” definitions and for the limitation of States’ exposure to international arbitration, but the effect of these provisions remains to be tested in future cases.

The Interpretation of “Investment” in Arbitration Practice

The notion of “investment” has caused many controversies in investment arbitration and the interpretation has appeared more complicated in the ICSID arbitration. Article 25 (1) of the ICSID Convention stipulates that the Centre’s jurisdiction “shall extend to any legal dispute arising directly out of relation to an investment, between a Contracting State and a national of another Contracting State, which the Parties to the dispute consent in writing to submit to the Centre”. However, the ICSID Convention does not further define what constitutes an “investment”.⁴¹ Consequently, arbitral tribunals have adopted divergent approaches of interpretations on the term.

Divergent Interpretations in ICSID Arbitrations

Starting with the Fedax case,⁴² many tribunals appear to have applied “objective” criteria or elements in assessing the existence of an investment under the ICSID Convention.⁴³ In the landmark Salini v Morroco, the tribunal confirmed that its jurisdiction depended on “the existence of an investment within the meaning of both the Bilateral Treaty as well as that of the Convention”, namely, a “dual test”.⁴⁴ More importantly, the tribunal held that “it would be inaccurate to consider that the requirement that a dispute be ‘in direct relation to an investment’ is diluted by the consent of the Contracting Parties. To the contrary, ICSID case law and scholarly writings agree that the “investment” requirement must be respected as an objective condition of the jurisdiction of the Centre”.⁴⁵ Accordingly, the requirements ratione materiae in the ICSID Convention must be satisfied for the purpose of ICSID jurisdiction, even though the dispute falls within the scope of the consent to arbitration by parties in investment treaties. Ultimately, the Salini tribunal set out four objective “elements” of investment for the purpose of ICSID jurisdiction, including the contribution, a certain duration, a participation in the risks, and the contribution to the economic development of the host State, which later became the famous “Salini test”.⁴⁶

Subsequently, many tribunals have adopted the Salini approach to articulate objective criteria for the notion of “investment” under the ICSID Convention and concluded that such elements should not be set aside by a consent of parties in investment treaties.⁴⁷ In the Patrick Michell v Congo annulment, for example, the ad hoc Committee held that:

The parties to an agreement and the States which conclude an investment treaty cannot open the jurisdiction of the Centre to any operation they might arbitrarily qualify as an investment. It is thus repeated that, before ICSID arbitral tribunals, the Washington Convention has supremacy over an agreement between the parties or a BIT.⁴⁸

Similarly, the Phoenix tribunal affirmed that:

“At the outset, it should be noted that BITs… cannot contradict the definition of the ICSID Convention. In other words, they can confirm the ICSID notion or restrict it, but they cannot expand it in order to have access to ICSID. A definition includes in a BIT being based on a test agreed between two States cannot set aside the definition of the ICSD Convention, which is a multilateral agreement. As long as it fits within the ICSID notion, the BIT definition is acceptable, it is not if it falls outside of such definition”.⁴⁹

Contrary to the Salini approach, some tribunals have advocated an “subjective” or the so-called “party autonomy” approach to understand the notion of “investment” under the ICSID Convention. In Biwater Gauff v Tanzania, the tribunal held that the Salini criteria were “not fixed or mandatory as a matter of law”, and that according to the travaux préparatoires of the Convention, “the term was left intentionally undefined, with the expectation (inter alia) that a definition could be the subject of agreement as between Contracting States”.⁵⁰ According to the tribunal, the Salini test itself is “problematic” if the identified “typical characteristics” of an investment were elevated into “a fixed and inflexible test” which would result in “arbitrary exclusion of certain types of transaction” and contradictory definition to investment agreements.⁵¹ Ultimately, the tribunal held that “a more flexible and pragmatic approach to the meaning of ‘investment’ is appropriate, which takes into account the features identified in Salini, but along with all the circumstances of the case, including the nature of the instrument containing the relevant consent to ICSID”.⁵²

In the MHS v Malaysia Annulment decision, the ad hoc committee pointed out that “it is important to note that the travaux préparatoires do not support the imposition of ‘outer limits’ such as those imposed by the Sole Arbitrator in this case”.⁵³ Furthermore, the annulment committee highlighted the importance of the 2800 investment treaties signed after the adoption of the ICSID Convention) and held that: “[i]t is those bilateral and multilateral treaties which today are the engine of ICSID’s effective jurisdiction. To ignore or depreciate the importance of the jurisdiction they bestow upon ICSID, and rather to embroider upon questionable interpretations of the term ‘investment’ as found in Article 25(1) of the Convention, risks crippling the institution”.⁵⁴

Furthermore, the Inmaris tribunal pointed out that:

…it will be appropriate to defer to the State’s parties’ articulation in the instrument of consent (e.g. the BIT) of what constitutes an investment. The State parties to a BIT agree to protect certain kinds of economic activity, and when they provide that disputes between investors and States relating to that activity may be resolved through, inter alia, ICSID arbitration, that means that they believe that that activity constitutes an ‘investment’ within the meaning of the ICSID Convention as well. That judgment, by States that are both Parties to the BIT and Contracting States to the ICSID Convention, should be given considerable weight and deference. A tribunal would have to have compelling reasons to disregard such a mutually agreed definition of investment.⁵⁵

As outlined above, a major difference between the “Salini approach” and the “party autonomy” approach to interpret Article 25 of the ICSID Convention is whether the term “investment” under Article 25 contains objective/inherent meaning that places a limit on State parties’ definition of “investment” in other legal instruments such as BITs for the purpose of ICSID jurisdiction. Such a divergence carries on in recent arbitral decisions.

In Vestey v. Venezuela, the majority of the tribunal generally followed the Salini approach and held that “the term ‘investment’ in Article 25 of the ICSID Convention has an independent meaning… [and] comprises three components: a commitment or allocation of resources, risk and duration”.⁵⁶

In Quiborax v Bolivia, the tribunal held that “the ICSID Convention contains an objective definition of ‘investment’, which must be met regardless of whether that same test is also inherent to the term ‘investment’ used in the BIT or whether it is additional to the BIT definition”.⁵⁷ In the tribunal’s view, “the Contracting States to the ICSID Convention intended to give the term ‘investment’ an ‘ordinary meaning’ as opposed to a ‘special meaning’.⁵⁸ The tribunal further pointed out that “investor-State cases indeed given substance and content to an objective meaning of ‘investment’.⁵⁹ Therefore, “[w]hether the objective test under the ICSID Convention is independent from and additional to the definition found in the BIT, or whether the same objective test is inherent to the term investment used in the BIT, the Tribunal must in any event review the elements of the objective definition to ascertain the existence of an investment”.⁶⁰

In Philip Morris v Uruguay, the tribunal held that the concept of “investment” is “central to the Centre’s jurisdiction and the Tribunal’s competence ‘ratione materiae’”.⁶¹ More importantly, the tribunal pointed out that the consent of the Contracting Parties under the BIT to the scope of “investment” is of relevance when establishing the meaning of the term under Article 25(1) of the ICSID Convention, though such Parties do not have an unfettered discretion to go beyond what have been called the “outer limits” set by the ICSID Convention) , the establishment of which be based on “the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose”.⁶²

In Orascom v. Algeria, the tribunal considered that the meaning of “investment” under the ICSID Convention was an objective one, which included elements of (a) a contribution or allocation of resources; (ii) a duration; and (iii) risk, which included the expectation (albeit not necessarily fulfilled) of a commercial return.⁶³ The tribunal agreed with the Saba Fakes award that these requirements “are both necessary and sufficient to define an investment” under the ICSID Convention .⁶⁴

In contrast, in OIEG v Venezuela, the tribunal held that:

States enjoy wide discretion to define which investments they wish to protect through a BIT and that Article 25 (1) of the ICSID Convention should not be subject to a restrictive interpretation. If two States have included a certain asset within a list of investments, a tribunal should only exclude it if it fails to meet the requirements of the objective and inherent concept of investment, if there is a compelling reason to do so.⁶⁵

Likewise, the Kim and others v. Uzbekistan tribunal held that to establish jurisdiction, “the claim must pass both through the institutional jurisdictional keyhole set forth in Article 25 as well as the specific jurisdictional keyhole defined in the BIT”; and the requirements of the ICSID Convention “were cognizant that the parties to a particular BIT may construct a more specific jurisdictional keyhole in their instrument”.⁶⁶

Salini Test: “Jurisdictional requirement” or “Characteristics/Indicia”?

The features or elements of “investment”, known as “Salini test”, has long been contested by investment arbitration community in the appreciation and application of the term. Despite of the divergence, most tribunals seems to accept that the term “investment” under the ICSID Convention contains certain features or inherent elements that can provide guidelines for identifying investment. As stated by Schreuer, “[it] would not be realistic to attempt yet another definition of ‘investment’ on the basis of ICSID’s experience. But it seems possible to identify certain features that are typical to most of the operations in question… These features should not necessarily be understood as jurisdictional requirements but merely as typical characteristics of investments under the Convention.”⁶⁷

In fact, what many tribunals have rejected seems to be rigid application of Salini test/criteria as jurisdictional requirements. In Phillip Morris v Uruguay, for instance, the tribunal opined that “there is no such a ‘jurisprudence constante’ with respect to acceptance of the Salini test”.⁶⁸ In the tribunal’s view, the four constitutive elements of the Salini test were merely “typical feature of investments under the ICSID” which might “assist in identifying or excluding in extreme cases the presence of an investment”, but they were not “a set of mandatory legal requirements” and could not “defeat the broad and flexible concept of investment under the ICSID Convention to the extent it is not limited by the relevant treaty”.⁶⁹

Indeed, the success of “Salini test” relies on that it “substantiated” the “ordinary meaning” of the term “investment” under the ICSID Convention by identifying certain constitutive elements. But its failure to receive wider acceptance was precisely due to the imposition of an extra-textual elements into the list of the constitutive elements, such as the contribution to the host States’ development.

In KT Asia v Kazakhstan, the tribunal held that “…[t]he absence of a definition of ‘investment’ under the ICSID Convention implies that the Contracting States intended to give to the term its ordinary meaning under Article 31 (1) of the VCLT as opposed to a special meaning under Article 31 (4) of the same treaty”.⁷⁰ In this connection, the constitutive elements of the “ordinary meaning” of “investment” need to be clarified and concentrated on the inherent hardcore aspects of the term, such as contribution, duration, risk and return expectations, as identified in arbitral decisions, but without further qualifications on other aspects such as “substantial contribution” or “regular return” or imposition of any extra-textual requirements such as legality, good faith or contribution to the host state’s development.

In Romak v. Uzbekistan, the tribunal resorted to the Black’s Law Dictionary that identified the “ordinary/inherent meaning” of the term “investment” to include “the commitment of funds or other assets with the purpose to receive a profit, or ‘return’, from that commitment of capital”.⁷¹ In Alpha v Ukraine, the tribunal was reluctant “to apply a test that seek to assess an investment’s contribution to a country’s economic development” that would “impose additional requirements beyond those expressed on the face of Article 25 (1) of the ICSID Convention”.⁷² The tribunal held that:

Should a tribunal find it necessary to check whether a transaction falls outside any reasonable understanding of ‘investment’, the criteria of resources, duration, and risk would seem fully to serve that objective. The contribution-to-development criterion, on the other hand, would appear instead to reflect the consequences of the other criteria and brings little independent content to the inquiry. At the same time, the criterion invites a tribunal to engage in a post hoc evaluable of the business, economic, financial and/or policy assessment that prompted the claimant’s activities. It would not be appropriate for such a form of second-guessing to drive a tribunal’s jurisdictional analysis.⁷³

The above conforms with the “ordinary” meaning of the term “investment” as for example defined in the Oxford English Dictionary (OED), namely, “the act of investing money in something”, and “invest” is defined as “to buy property, shares in a company, etc., in the hope of making a profit”. Clearly, the most fundamental elements involved under the OED definition of “investment” are “contribution” (i.e., money or other resources “to buy property, shares in company etc.”) and “return expectation” (i.e., “hope of making a profit”). And from the “return expectation” (i.e., “hope of making a profit”), the “risk” and “duration” elements can also deduced. Notably, the “ordinary” meaning of “investment” illustrated in the definition under OED does not require the contribution to be “substantial”, or the expected profit or return to be “regular”, let alone any reference to extra-textual aspects such as legality, good faith or (significant) contribution to the host state’s development.

Indeed, if the constitutive elements of the “investment” were to be so clarified to include only the minimum, inherent core aspects of the term, there should be little hesitance for tribunals to apply such “ordinary meaning” and its constitutive elements as jurisdictional requirements, as there should be little difference in the result adopting either of the “jurisdictional requirement” or “characteristics/indicia” approach. The GEA tribunal seems to accept this position by noting that whatever test was applied, each lead to the same conclusion.⁷⁴

A “Double Door” Theory

The “dual test” or “double keyhole” theory has been generally accepted in case of ICSID arbitration, requiring the “investment” in question to meet both the requirement under Article 25 of the ICSID Convention and definition of “investment” under the applicable investment treaties. With respect to the ICSID requirement, it is necessary to recall the Report of the World Bank’s Executive Directors on the ICSID Convention:

While consent of the parties is an essential prerequisite for the jurisdiction of the Centre, consent alone will not suffice to bring a dispute within its jurisdiction. In keeping with the purpose of the Convention, the jurisdiction of the Centre is further limited by reference to the nature of the dispute and the parties thereto.…

No attempt was made to define the term ‘investment’ given the essential requirement of consent by the parties, and the mechanism through which Contracting States can make known in advance, if they so desire, the classes of disputes which they would or would not consider submitting to the Centre.⁷⁵

According to the first paragraph, it is important to note that jurisdictional requirements under the ICSID Convention are independent from the jurisdictional requirements under the applicable investment treaties providing the consent to ICSID arbitration. While the parties’ consent is critical for ISDS, it alone does not necessarily fulfil jurisdictional requirements for ICSID arbitration, as certain “outer/further limits” are imposed by the ICSID Convention.

Therefore, a “double door” theory might be more accurate than the “dual test” or “double keyhole” theory as the claim has to pass two “doors” to fulfil the ICSID jurisdictional requirements: the first door refers to the jurisdictional threshold set by the applicable investment treaty in terms of investor-state dispute settlement; and the second door refers to the jurisdictional threshold set by Article 25 of the ICSID Convention only in the case of ICSID arbitration. This has been confirmed by numerous investment arbitration decisions as well as rejections by the ICSID Secretary General to register certain disputes (such as pure commercial sales disputes) for ICSID arbitrations.

In this connection, the “ordinary meaning” of investment should apply in the interpretation of Article 25 of the ICSID Convention in accordance with Article 31 (1) of the VCLT. In fact, even tribunals that rejected the “Salini approach” had accepted that “Salini test” could assist in including or excluding certain extreme transaction. In Alpha v Ukraine, while the tribunal ultimately deferred to the BIT definition, it also clarified that not “any definition of ‘investment’ that might be agreed by States in a BIT… must constitute an ‘investment’ for purpose of Article 25 (1). To cite the classic example, a simple contract for the sale of goods, without more, would not constitute an investment within the meaning of Article 25(1), even if a BIT or a contract defined it as one”.⁷⁶ Hence, “elements discussed in the Salini test might be of some use if a tribunal were concerned that a BIT or contract definition of ‘investment’ was overreaching and captured transactions that manifestly were not investment under any acceptable definition”.⁷⁷

The second paragraph of the Report of the Bank’s Executive Directors on the ICSID may support the above understanding. As noted by many tribunals and commentators, the undefined “investment” under Article 25 exhibits a balance achieved between capital exporting and importing States in the process of negotiating the Convention, as the former succeeded in securing a “no definition” of “investment” in the Convention imposing no extra qualifications on the term (such as a minimum amount of asset invested, or a debt of at least certain years as originally proposed), whilst the latter were satisfied by the possibility of excluding classes of disputes that they would not want to be submitted to the Centre.⁷⁸ Nonetheless, it was not really a “balance”, since whilst the former’s success was substantial and substantiated as demonstrated in numerous arbitration cases, the latter’s satisfaction was illusionary, as the said declarations turned out to be of merely declaratory value and could not amount to jurisdictional exclusions.⁷⁹ Such factual imbalance certainly enhances the justification for the application of the “ordinary meaning” of “investment” as a jurisdictional requirement for ICSID arbitrations. In this regard, the ordinary meaning constitutes implied terms of the relevant treaties, which shall apply unless it has been overridden by explicit treaty provisions conferring it another “special meaning”, as stipulated under Article 31 (4) of the VCLT.

“Inherent Meaning” of Investment in BITs

Another related questionable issue is whether or not the same “ordinary meaning” and accordingly its constitutive elements apply to the applicable investment treaties. According to some recent tribunals, the term “investment” per se contained an objective meaning, whether it is mentioned in the ICSID Convention or in a BIT.⁸⁰ In the Romak v Uzbekistan, the non-ICSID tribunal observed that “[t]he term ‘investment’ has a meaning in itself that cannot be ignored” when considering the definition clause under the BIT.⁸¹ The tribunal therefore held that:

…the term ‘investment’ under the BIT has an inherent meaning (irrespective of whether the investor resorts to ICSID or UNCITRAL arbitral proceedings) entailing a contribution that extends over a certain period of time and that involves some risk… By their nature, asset types enumerated in the BIT’s non-exhaustive list may exhibit these hallmarks. But if an asset does not correspond to the inherent definition of ‘investment’, the fact that it falls within one of the categories listed in Article 1 does not transform it into an ‘investment’.⁸²

In KT Asia v Kazakhstan, the tribunal held that an objective meaning not only existed under Article 25 of the ICSID Convention, but also within the definition of “investment” under the applicable BIT. As the tribunal noted, the absence of definition under the ICSID Convention implied to give the term “investment” ordinary meaning under Article 31 (1) of the VCLT as opposed to a special meaning under Article 31 (4) of the same treaty.⁸³ The objective meaning is inherent to the word “investment”, regardless of the application of the ICSID Convention.⁸⁴

Likewise, Orascom v Algeria tribunal considered that “objective” or “inherent” meaning was also present in a BIT’s definition of investment.⁸⁵ The use of the term “investment” in both the ICSID Convention and the BIT “imports the same basic economic attributes of an investment derived from the ordinary meaning of that term, which comprises a contribution or allocation of resources, duration, and risk”.⁸⁶

The Isolux tribunal adopted a similar position. The tribunal pointed out that the list of assets in the ECT definition “provides examples of investment but does not define the concept…”⁸⁷ The tribunal considered that the absence of a definition of investment was the common feature of bilateral and multilateral treaties which justified a definition of the concept of investment.⁸⁸ According to the evolution of arbitral jurisprudence, the tribunal noted that the objective definition of investment included a contribution, the receipt of returns and the assumption of risks.⁸⁹

In Masdar v Spain, the tribunal acknowledged that “a substantial number of recent investor-State awards have considered that the term ‘investment’ has an inherent meaning, which an alleged investment must meet in addition to falling into one of the categories of assets generally mentioned in BITs. Importantly, these awards have applied this so-called inherent, or objective, definition not only when applying the ICSID Convention, but also when interpreting BITs”.⁹⁰ In the tribunal’s view, “elucidating the meaning of the term ‘investment’ in Article 1 (6) of the ECT is part of the interpretation of that provision”.⁹¹

However, some tribunals adopted different views. The SGS tribunal suggested that “it would go too far to suggest that any definition of investment agreed by states in a BIT (or by a state and an investor in a contract) must constitute an ‘investment’ for purpose of Article 25(1)”.⁹² In A11Y v Czech, the non-ICSID tribunal noted that there were no definition or limitation in Art 1(a) of the BIT of the term which only refers to “every kind of asset belonging” to the investor without any further qualification.⁹³ Accordingly, the tribunal considered that “[t]he Contracting Parties to the Treaty could have qualified the definition of investment but they chose not to do so. It is [therefore] not the task of this Tribunal to add words to the broad definition agreed by the Contracting Parties”.⁹⁴

In this connection, whether or not the constitutive elements of objective/ordinary meaning of investment applies to the applicable investment treaties should depend on the wording of the relevant provisions. To be specific, if the governing treaty does not actually “define” the term “investment” or otherwise conferring a “special meaning” onto “investment”, but merely provides some illustrations (or indicia) of what should or should not be included under the term (even if it is placed in the “definition” provision), it would be appropriate and indeed necessary to import and apply the “ordinary meaning” and its constitutive elements to help out in treaty interpretation.

However, if the governing treaty has clearly “defined” the term “investment” to the extent that a “special meaning” under Article 31 (4) of the VCLT can be established as having been conferred onto the term, e.g., in the formula of “investment means…”, it would be inappropriate to impose the said “ordinary meaning” and elements to qualify such clear, explicit treaty definition, even if the treaty definition manifestly deviates from the “ordinary meaning”. In Yukos case, the tribunal recalled again that “according to Article 31 of the VCLT, a treaty is to be interpreted in good faith in accordance with the ordinary meaning of its terms” and thus read Article 1(6) of the ECT as containing the “widest possible definition”.⁹⁵ In Energoalians v Moldova, the tribunal accepted the point of view that “the ECT in its nature is a document that differs from other investment treaties and its rather extensive wording and a detailed list of what constitutes an investment does not allow to change or narrow down such a broad definition through an abstract and conceptual approach”.⁹⁶ In A11Y v Czech, the tribunal also held that the contracting parties “could have qualified the definition of investment but they chose not to do so”. Hence, the tribunal should not “add words to the broad definition agreed by the Contracting parties ”.⁹⁷

Conclusion

To conclude, the concept of “investment” as a key term and prerequisite for investor-State arbitration has caused many difficulties and controversies in ISDS practice. In particular, the silence on the term “investment” under the ICSID Convention has given rise to divided interpretations in arbitral decisions on whether the tribunal should appreciate and apply the “ordinary, objective or inherent’ meaning of the term “investment” to establish the ICSID jurisdiction. Many tribunals adhered to an “objective” approach or the so-call “Salini test” to identify several constitutive elements of “investment”, such as recourses commitment, certain duration, assumption of risk, and contribution to the development of the host State. However, other tribunals advocated a “subjective” or “party autonomy” approach to defer to the definition under the applicable substantive investment treaties. In recent arbitral cases, some tribunals have gone further to apply the “Salini test” to non-ICSID arbitration, which was rejected by other tribunals.

In response to the controversy and uncertainty in arbitral practice with respect to the broad asset-based definition of investment, many States have used various approaches to narrow and clarify the scope of investment for the avoidance of exposure to ISDS claims. Three features are prominent in recent investment practice, namely, the inclusion of the characteristics of investment, the explicit exclusion of certain assets, and the requirement of in accordance with host state’s laws. The development in investment treaties on the definition of investment reflects States’ efforts to strike a balance between attracting and protecting foreign investment and safeguarding state’s right to regulate for legitimate public policy objectives. Nonetheless, the effect of these newer investment treaties remains to be tested in arbitral cases.

For future ISDS interpretations of the concept of “investment”, a “double door” theory may be useful to address the relationship between the Article 25 of the ICSID Convention and the definition of “investment” clause under application investment treaties. For the ICSID jurisdiction, the “objective and inherent meaning” of the term “investment” should be appreciated and applied in accordance with Article 31(1) of the VCLT, which include only the minimum hardcore of the concept. The same ordinary meaning might also be applied in non-ICSID arbitrations only if the governing substantive treaty does not clearly define what constitutes an “investment”. In this regard, the definition of investment clause under investment treaties is likely to subject to greater scrutiny when tribunals apply the “ordinary meaning” of investment in case the substantive investment treaty failed to provide a “special meaning” under Article 31(4) of the VCLT.

Cross-References

• Jurisdictional Objections and Defenses (Ratione Personae, Ratione Materiae, and Ratione Temporis)

• Protection of Cross-Border Data Flows Under International Investment Law

• The Definition of Investor in Investment Treaty Arbitration