The debate surrounding the meaning and scope of the term ‘investment’ under the ICSID Convention is a product of the larger tussle between capital exporting and capital importing states, which convened at Washington in the search for a mutually beneficial agreement on foreign investments. It has been argued by Prof. Julian Davis Mortenson that ‘investment’ under Article 25 of the ICSID Convention must be interpreted as broadly as possible, in deference to party autonomy. An analysis of the travaux préparatoires leads him to conclude that parties’ consent should be the sole basis for evaluating whether a particular enterprise can be classified as an investment. Prof. Mortenson’s previous posts on this topic on the blog can be found here, here and here. While agreeing with such an approach in principle, this post demonstrates how it is not an effective alternative in practice. Instead, I explore the existence of a reasonable middle ground that allows the intentions of parties to be given effect to, while keeping the core of the Convention intact.


Contextualising State Consent

The avowed object of the ICSID Convention is to encourage the flow of private international investment by facilitating the settlement of disputes between investors and host states. The need to maintain a ‘careful balance’ between the interests of investors and the host states is, therefore, stated to be an important task of the Convention’s provisions.

Prof. Mortenson notes that it is this balancing act that resulted in the term ‘investment’ deliberately being left undefined. Such a ‘compromise’, first proposed by the United Kingdom, is espoused as a consent-based approach to jurisdiction, which offers sufficient flexibility to contracting parties to delimit the scope of covered investments. States are free to have their own definitions of what constitutes an ‘investment’ either through subsequent Bilateral Investment Treaties (BITs) or as unilateral notifications to the ICSID Secretariat (referred to as ‘opt-outs’). However, it is important to keep in mind that this was not an intentional choice, but one made out of compulsion. It is only because contracting states were unable to reach a consensus that the term was left open-ended. Moreover, whether or not such a choice has, in fact, ensured flexibility and encouraged state autonomy is questionable. For instance, only a handful of states have employed notifications under Article 25(4) to exclude certain classes of disputes from ICSID jurisdiction, and even such notifications have been held to be merely informational and not binding qualifications on state consent. Further, limiting the scope of states’ consent through arbitration agreements in BITs or specific contracts has also not been widely embraced insofar as most BITs continue to incorporate an expansive asset-based definition of investments, with reference to ICSID arbitration for the purposes of dispute settlement.

Indeed, if one were to agree that the compromise formula based on party autonomy truly gives effect to the consent of all states, state practice would lead to the inevitable inference that host states have come to accept the all-encompassing definition of ‘investment’. However, that is understandably not the case. Moreover, the expression of consent requires a positive act. Merely because states have not chosen to limit their consent to a specific class of investments through such opt-out mechanisms can not, by itself, be taken to imply an acceptance of the default standard without limitations. Such an approach must also be considered in light of prevailing geopolitics, acknowledging that some states possess greater bargaining powers than others.


Requirement of an Outer-Limit

Article 25 of the ICSID Convention explicitly limits jurisdiction ratione materiae to ‘any legal dispute arising directly out of an investment’. The deferential approach espoused by Prof. Mortenson requires that the determination of jurisdiction be solely based on parties’ conception of what constitutes investment, as found in BITs or other instruments of consent, subject only to the exclusion of ‘facially absurd’ non-economic claims. In other words, the term ‘investment’ under Article 25 is seen as non-justiciable and lacking a real limiting effect.

On the other hand, while acknowledging that consent is a jurisdictional prerequisite for bringing claims under the ICSID Convention, the Report of the Executive Directors states that ‘consent alone will not suffice to bring a dispute within its jurisdiction.’ In other words, mere consent cannot confer jurisdiction where none otherwise exists. A similar inference that parties’ agreement does not trump all other requirements can also arguably be drawn from Rule 41 of the ICSID Arbitration Rules, which empowers a tribunal to consider questions of jurisdiction ‘on its own initiative’. Interestingly, at the negotiations for drafting the ICSID Convention, there was a proposal to do away with the term ‘investment’ under Article 25 entirely. The same was however rejected, evidencing that the term has a meaning and effect which cannot be sidestepped by the mere consent of parties.

Accordingly, it has been suggested that the very act of consenting to ICSID arbitrations should be taken to reflect state parties’ intention to ‘overlay the requirements of the ICSID Convention over the broad BIT definition of investment’. Moreover, if consent alone was the jurisdictional requirement, then given the relative certainty of enforcement, all private contracts entered into by states with foreign parties would adopt ICSID arbitrations, effectively rendering all other arbitral mechanisms redundant.

As noted by Prof. Schreuer, the suggestion that the term ‘investment’ under Article 25 extends to all plausible economic activity as consented to under the BITs is also at odds with the well-accepted idea that purely commercial transactions do not fall within the scope of investment disputes. The very fact that there exists a threshold beyond which an activity or enterprise will not be characterised as an investment, points to an effective outer-limit implicit in the understanding of the term itself. It is for the same reason that in 1999, the Secretary-General of ICSID refused to register an arbitration case on the ground that the transaction therein was not capable of being classified as an investment.


Scope of the Article 25 Prerequisite

If the view of absolute deference to party autonomy is an extreme standpoint, the restrictive four-prong test expounded in Salini v. Morocco represents the other extreme. It effectively converts typical descriptive characteristics of investments into rigid, legally binding pre-requisites necessary to establish jurisdiction. Instead of qualifying outer-limits on arbitral jurisdiction in investment disputes, the Salini test postulates an independent and exhaustive appraisal of the term ‘investment’, requiring strict objective compliance. In doing so, it pays scant regard to parties’ agreements defining investments, by reason of which it has become a particularly divisive ruling.

Despite the acceptance of the Salini jurisprudence in subsequent decisions, some arbitral tribunals have differed and adopted approaches, which, if taken together, may be able to harmonise party autonomy with the scope of the ICSID Convention. For instance, it has been held that a prior agreement of parties with respect to what constitutes an investment creates a strong presumption in favour of jurisdiction under Article 25 as well. It has also been recognised in various rulings that a jurisdictional requirement cannot be imposed in terms of a strict, objective test, the yardsticks of which must only be seen as mere examples illustrative of typical investments, and nothing more. At the same time, the requirement to keep manifestly non-investment disputes out of the purview of ICSID arbitrations has also been acknowledged, in which case the flexible application of objective criteria in the context of particular facts and circumstances may be useful.

The underlying symbiotic relationship between the ICSID Convention and BIT provisions is central to the reconciliation of two distinct approaches arguing for the primacy of one instrument over the other. It has been proposed that the term ‘investment’ must be defined ‘in the context of shared systemic objectives underlying the mechanism of investment protection’. In this regard, what Prof. Emmanuel Gaillard has referred to as the ‘intuitive school of thought’ provides an insightful basis for a resolution. It places emphasis on identifying, rather than defining, characteristics of investments by means of guiding factors, with significant room for subjective discretion. Such an approach allows a case-by-case evaluation of investments based on criteria not limited to the four prongs of Salini. By not insisting on their cumulative compliance, but on an inclusive appraisal of all facts and circumstances, it may potentially lead to a decision that is flexible, less restrictive and perhaps more in line with parties’ intentions as well.



The clash of the deferential and restrictive approaches is a quintessential embodiment of some of the central issues in the practice of international law, viz. state consent and the exclusion of stare decisis. It is evident, however, that state parties share conflicting, but complementary interests in giving effect to the provisions of the ICSID Convention. An approach allowing an initial presumption in favour of party autonomy, subject to a flexible inquiry based on the ordinary meaning of the term ‘investment’, seems to be well suited to reasonably balance competing interests. The resulting degree of uniformity in decision making may allow for greater certainty in parties’ expectations, leading to better outcomes in line with the stated objects of the ICSID Convention itself.


To make sure you do not miss out on regular updates from the Kluwer Arbitration Blog, please subscribe here. To submit a proposal for a blog post, please consult our Editorial Guidelines.

Profile Navigator and Relationship Indicator
Offers 6,200+ data-driven arbitrator, expert witness and counsel profiles and the ability to explore relationships of 13,500+ arbitration practitioners and experts for potential conflicts of interest.

Learn how Kluwer Arbitration Practice Plus can support you.

Kluwer Arbitration
This page as PDF

One comment

  1. Doesn’t the scope of the term investment derive from Article I of the Articles of Agreement of the World Bank? There it seems not to be limited in meaning, but in terms of the constitutional purposes of the bank has nuances like – it contributes to peace and development, has a productive purpose – raising productivity, standard of living and labor conditions in the territory of the member state, and in those senses is useful and urgently needed.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.