For some time now, the world has seen a rise in proto-nationalism, protectionism, and even nationalization of resources. This paradigm shift when coupled with criticism being leveled against the Investor-State Dispute Settlement (“ISDS”) system for its alleged bias in favor of capital exporting countries, make for a dangerous combination. With the ‘return of the state’ in ISDS (most recently evidenced by  India’s new BIT with Brazil) International Centre for Settlement of Investment Disputes (“ICSID”) arbitrations have never been more vulnerable to the threat of denunciation (particularly in those investment agreements which do not have a fallback to arbitration under the UNCITRAL or other rules).

It is in this context that the author wishes to briefly recall the law governing denunciation under the Convention – particularly Mr. Christer Söderlund’s separate opinion in Blue Bank International (“Opinion”). In the ocean of literature available on this topic the author believes the Opinion to be insufficiently explored. Although not law, it presents a very interesting thesis departing from the conventional wisdom.

Under the existing framework, denunciation is governed by Article 71 and Article 72 of the Convention. The proper interpretation of the articles has been the subject of widespread academic discussion and even some ICSID awards. For example, it has been discussed several times on this blog, e.g. here, here, here & here. Broadly speaking, pursuant to Article 71 a contracting state may denounce the Convention by furnishing a notice to the World Bank. The denunciation will take effect six months after the notice. Notwithstanding this, Article 72 provides that the aforesaid notice will not affect any rights or obligations ‘arising out of consent to the jurisdiction of the Centre given by one of them before such notice was received by the depositary’. This opens up a key question: what is the effective date of denunciation taking effect?

Three schools of thought have emerged. The first school advocates that denunciation takes effect forthwith, i.e. immediately after the notice is received by the World Bank. The most notable advocate for this school has been Professor Schreuer, whose views were heavily relied upon in the 2018 Fabrica award. The second school of thought advocates that denunciation takes effect six months after the notice. The result of this view would be that investors would have up to six months to commence ICSID arbitration if the denouncing state has offered such procedural redress in an applicable instrument. This view was adopted by the Tribunals in Blue Bank International and Venoklim. The third school postulates that investors can invoke the arbitration clause even after the six-month period. The result of this view is that investors would be able to invoke the right to arbitrate at any point up to the expiry of the sunset clauses (if any) in the consent instrument.

Söderlund’s Opinion appears to fall within the third school of thought. What is indeed remarkable is that an attempt has been made to challenge the working contractual model based on which other tribunals (including the majority in Blue Bank International) have rendered their decisions, rather than suggesting an alternative interpretation under the already popular framework.


The offer-acceptance model

Generally speaking, an arbitration clause provided for in a treaty is considered to be based on the offer-acceptance model. In other words, it is a “standing offer” by the host states which may be unilaterally accepted by Investors.

Advocates of the first school of thought argue that denunciation of the Convention in effect revokes this standing offer. This view is underpinned by an interpretation of Article 72 in light of Article 25 of the ICSID Convention. Article 25(1) provides that: ‘..when the parties have given their consent, no party may withdraw its consent unilaterally’. Therefore, any obligation which may subsist by the operation of Article 72 is qualified by the requirement of consent of both the parties (“Perfected Consent”). In other words, on a contextual interpretation of Article 72, both parties are required to honor their existing obligations unless mutually agreed otherwise by the parties. Since the Investor has not accepted the standing offer, all obligations of the host state stand extinguished. It is further argued that Article 71 has no place in this discussion. This is because Article 71 governs the rights/obligations of contracting states as party to the Convention, whereas Article 72 governs the rights/obligations of Contracting states as a party to arbitrations. Additionally, the cooling off period in Article 71 is applicable only to those right/obligations which do not require Perfected Consent. For example, the right to participate in the Administrative Council (Article 4-7 of the Convention), the duty to respect the immunities and privileges foreseen in the Convention (Articles 18–24) do not require Perfected Consent (a unilateral undertaking by the state is sufficient) and are therefore governed by Article 71. In other words, Article 71 encompasses a general rule whereas Article 72 is a specific rule with regard to rights/obligations arising out of consent. Since it is now a well-established principle of interpretation that a specific rule supersedes a general rule (lex specialis derogat legi generali), it is only proper that denunciation takes effect forthwith.

Advocates of the second school also start their enquiry under the guise of the offer-acceptance framework, but nevertheless reach a somewhat different conclusion. Adopting a purposive interpretation, it is argued that there exists no dichotomy between Article 71 and 72 and any other view would be at the cost of “robbing” the cooling off period in Article 71 of its effect utile. This view is in consonance with International treaty practice (see e.g. Article 56(2) of The Vienna Convention on the Law of Treaties).

Some scholars have made arguments which can be classified under the third school (For example- Professor Gaillard seems to offer something on the lines of a firm offer model).. Briefly, he demarcates those arbitration clauses which purport a mere offer to arbitrate as against those clauses which incorporate an unconditional consent to arbitration. In other words, the nature of state consent and its possible revocation thereof varies according to the terminology used in the consent instrument.


Third party beneficiaries’ model

Mr. Söderlund’s separate opinion has challenged the offer-acceptance model when consent is expressed through an investment treaty such as a BIT. He does not dispute that Article 72 requires Perfected Consent; what he in fact argues is that this requirement has already been met. Instead, he challenges one major assumption in the previous model by presenting an alternative theoretical view-point, namely, that the investor is not the appropriate party who has the obligation to perfect consent because it is only a beneficiary to the contract/treaty. The real parties to the “contract” (i.e. the treaty) are the states and not investors.

When the Convention was formulated, state consent in relation to foreign investment typically featured in an arbitration clause entered into by the state itself – usually in a natural resource concession or an investment agreement (¶21). Given the historical account, the ensuing practice of including ISDS clauses in investment treaties was not clearly envisioned and thoroughly discussed by the drafters. The first Investment treaty to expressly incorporate an investor-state arbitration clause was the 1968 BIT between Indonesia & the Netherlands. Since then, over 2000 such BITs have entered into force. As Söderlund notes, this shift in practice means that  ‘the provisions of the Convention will nowadays have to be regularly applied in a contextual framework that did not exist at the time of the Convention’s creation’ (¶24).

Flowing from the above analysis, there may be two different but related reasons as to why the offer-acceptance model is logically inaccurate in a typical BIT context:

  1. In the case of an investment contract or an agreement to submit dispute to the International Court of Justice, the parties to the dispute are the same. However, in the case of an investment treaty the contracting parties are the host state and the home state of the investor, whereas the disputing parties are the host state and the investor.
  2. There are three relevant parties as opposed to only two parties. The offer-acceptance model ceases to make sense in a tripartite context as it is only applicable in a two-dimensional playing field.

The preferred model should, therefore, be the third-party beneficiaries’ model presented by Mr. Söderlund in his Separate Opinion. Put simply, the “contract” to arbitrate is between the two sovereign states, whereas the investor is a third-party beneficiary to the contract.

Consequently, the arbitration clause as contained in an investment treaty should not be considered as a “standing offer”, which has to be accepted by an investor, but rather it should be treated as a “procedural right” that is secured to the third-party beneficiary. Conclusively, this right is secured with the investor as long as the investment treaty is in effect (¶48,49). In the situation where the denouncing state terminates the relevant BIT also, the right will exist until the sunset period runs its course. Since it is a well-established principle of law that a right conferred upon a beneficiary cannot be altered without the consent of that beneficiary, denunciation will not have any effect because here the beneficiary i.e. the investor has not consented to the alteration of its rights.



Mr. Söderlund offers a very interesting perspective which has so far featured only in academic discussions. However, the Fábrica award which was passed a year after the Opinion discusses Article 72 in great detail and disagrees with Mr Söderlund’s view. On this note it is important to recollect that there is no strict doctrine of stare decisis in ISDS proceedings. The future is still open to tribunals to adopt a reasoning similar to that of Mr. Söderlund – off-setting the possible use of denunciation in such vulnerable times.


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.

Kluwer Arbitration
This page as PDF

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.