Junefield Gold Investments Limited v. The Republic of Ecuador. PCA Case No. 2023-35

The arbitral tribunal in Junefield v. Ecuador (Junefield) addressed a long-standing and increasingly prevalent issue in international investment arbitration—namely, the scope of jurisdiction under a restrictive investor-state dispute settlement (ISDS) clause. In its award, the tribunal majority adopted a broad interpretation of Article 9.3 of the China-Ecuador Bilateral Investment Treaty (BIT), which appears to limit arbitration to disputes concerning the amount of compensation for expropriation that is not settled after six months of negotiations. Based on this broad interpretation, the tribunal held that it had jurisdiction over the claimant’s expropriation claims. This expansive reading of a seemingly narrow dispute settlement provision underscores the ongoing interpretative divergence relating to restrictive ISDS clauses, particularly in cases involving indirect expropriation.

The Junefield award needs to be situated within the broader context of China’s early and conservative BIT-making practice. A significant number of China’s early-generation BITs (approximately 64 percent) contain similarly restrictive ISDS clauses that limit arbitral jurisdiction to “disputes ‘involving the amount of compensation for expropriation.’”¹ These provisions have proven particularly challenging in arbitral practice. In nearly half of the known China-related ISDS cases (fourteen out of twenty-nine), claimants have encountered jurisdictional hurdles arising from such narrowly framed clauses, which may exclude preliminary questions concerning the existence and legality of an indirect expropriation, as well as other substantive treaty breaches, including violation of fair and equitable treatment or the obligation to provide full protection and security, etc. The tribunals’ decisions remain divided: some have adopted broad interpretations and affirmed jurisdiction,² while others have applied narrow readings and declined jurisdiction altogether.³ It is within this bifurcated and unsettled jurisprudential landscape that the Junefield award arrives. It represents a further and significant development in the interpretation of restrictive ISDS clauses, as the tribunal endorsed a broad reading by introducing new interpretive perspectives. First, the tribunal majority emphasized that the basic treaty makes no distinction between the existence, legality, and quantum of an expropriation, thereby implying that all such aspects may fall within the tribunal’s jurisdiction (para. 252). Second, it reasoned that when a restrictive ISDS clause is read together with a Fork-in-the-Road clause, the disputes it covers must be treated as “self-sufficient,” encompassing the full range of expropriation-related issues, including existence, legality and compensation (para. 231). In addition, the award is also likely to influence at least four ongoing ISDS cases involving China and Chinese investors,⁴ in which tribunals are grappling with similar jurisdictional questions under restrictive treaty language.

* * * *

The dispute arose out of Junefield Gold Investment Limited’s (Junefield) investment in a gold and silver mining project in Azuay Province, Ecuador. Between 2013 and 2017, Junefield carried out extraction and related activities through its subsidiary, Ecuagoldmining, which held mining concessions rights in Ecuador covering three areas: Canoas, San Luis A2 and Miguir. In 2018, unidentified individuals broke into the project site where employees lived and worked. Although Ecuador deployed police and military forces, they were later withdrawn. By 2019, Junefield lost access to the concession areas. Concurrently, a group of individuals alleging adverse impacts from the mining project, along with several Indigenous organizations, initiated legal actions in Ecuadorian courts, leading to the suspension of mining activities and demilitarization of the site (paras. 56–63). Junefield claimed that these acts collectively amounted to expropriation and violated Fair and Equitable Treatment (FET) and Protection and Security (PS) under the China-Ecuador BIT.⁵ On Oct. 4, 2022, Junefield filed for arbitration before the Permanent Court of Arbitration (PCA) under the 2010 version of the Arbitration Rules of the United Nations Commission on International Trade Law (UNCITRAL Rules), pursuant to Article 9 of the China-Ecuador BIT.

The core issue in Junefield was whether the tribunal had jurisdiction not only over the amount of compensation for expropriation but also over the legality of expropriation and possible violations of other substantive treaty standards. Ecuador challenged the tribunal’s jurisdiction based on the narrowly framed language of Article 9.3 of the China-Ecuador BIT⁶ on four main grounds: (1) Article 9.3, properly interpreted under the Vienna Convention on the Law of Treaties (VCLT), limited jurisdiction to quantum-related disputes; (2) disputes concerning the legality of expropriation or other substantive breaches were to be resolved by domestic courts; (3) the fork-in-the-road (FIR) clause did not support broader jurisdiction; and (4) the Most-Favored-Nation (MFN) clause could not expand jurisdiction beyond the limits set out in the Treaty (paras. 69–70). However, the tribunal majority rejected most of these objections and held that Article 9.3 was broad enough to cover both the legality of the claimed expropriation and the amount of compensation owed. One arbitrator dissented, arguing that the majority’s interpretation “significantly expanded the scope of the tribunal’s jurisdiction beyond that which the drafters of the Treaty intended.”⁷ Despite this divergence, all three arbitrators agreed that the tribunal lacked jurisdiction over other non-expropriation claims, as Article 9.3 of the China-Ecuador BIT expressly limited arbitral jurisdiction to dispute concerning the amount of compensation for expropriation, and that the MFN clause could not broaden Ecuador’s consent to arbitration.

Jurisdictional Limits Under Restrictive ISDS Clause: Quantum or Beyond?

To address this core jurisdictional issue, the tribunal began its analysis with a close, plain-language reading of Article 9 of the China-Ecuador BIT, focusing in particular on Article 9.3. It observed that Article 9.1 established the requirement of amicable negotiations as a precondition for adjudication, and that Article 9.2 then allows any dispute, if not resolved amicably, to be submitted to the host state’s local courts. Article 9.3, however, created a more specific pathway for arbitration: it permitted disputes “involving the amount of compensation for expropriation” to be submitted to an ad hoc tribunal after the same six-month period, provided the investor had not already submitted the case to domestic courts.⁸ The tribunal reasoned this concurrent timing—where both arbitration and local court proceedings could commence after the six months—reflected a Fork-in-the-Road mechanism, where choosing one forum excluded the other. This legal framework, it held, made the type of disputes contemplated by Article 9.3 as self-sufficient (emphasis added) and did not require prior domestic adjudication on the existence or legality of expropriation (para. 231).⁹

Accordingly, the tribunal rejected Ecuador’s argument that recourse to arbitration would be available only after the “proclamation” of an expropriation, that is, after a formal act of expropriation had been issued either by an administrative authority or confirmed by a domestic court. The tribunal explained that Ecuador’s argument could, in theory, make sense only in the case of direct expropriation, where a formal administrative decision declaring expropriation would already exist, leaving only the amount of compensation to be determined. However, this reasoning could not apply to indirect expropriation, which, by its nature, involves no such formal act and thus requires a determination, in the first instance, of whether an expropriation has occurred and, if so, an assessment of the corresponding compensation (paras. 233–34).¹⁰ Since Article 9.3 made no distinction between direct and indirect expropriation, the tribunal considered that requiring a prior domestic ruling would contradict the plain text of the Treaty. To support this, it further examined Article 4, which defined lawful expropriation, and found no requirement for domestic court to determine the occurrence or legality of an expropriation. Unlike other BITs (e.g. China-Singapore BIT, China Ghana BIT), which explicitly include such a requirement, the relevant language is absent in the present Treaty (paras. 236–37). As a further blow to Ecuador’s position, the tribunal also noted that the Treaty made no distinction among the existence, legality and quantum (emphasis added) of an expropriation, indicating that all aspects could fall within the jurisdiction.¹¹

Turning to the interpretive core, the tribunal analyzed the phrase “involving the amount of compensation for expropriation” under Article 31 of the VCLT as covering not only compensation but also the prior question of whether an expropriation had occurred. This reading was particularly relevant in the scenario of indirect expropriation, where no official declaration might exist, and thus the tribunal must itself determine the existence of expropriation to resolve the compensation issue (para. 241). The tribunal further stressed that the term “involving” could not be ignored or treated as neutral, rather, dictionary definitions confirmed its inclusive character, and this interpretation gave full effect to all terms in the provision without inserting restrictive conditions not found in the text. It also noted that this reading remained consistent with the FIR clause in the final sentence of Article 9.3 (para. 243).¹²

Although the tribunal acknowledged Ecuador’s reliance on precedents and comparator BITs, it ultimately gave them limited weight (paras. 250, 257). The tribunal noted that the wording of other “first generation” Chinese BITs often differed significantly from the present Treaty and frequently included express domestic review requirements or lacked FIR clauses.¹³ Among cases with identical or nearly identical treaty language, arbitral decisions had also been divided.¹⁴ In this context, the tribunal gave greater weight to cases like Sanum v. Laos, where similarly worded clauses were interpreted as allowing tribunals to assess both expropriation existence and compensation (para. 259). Ecuador further invoked China’s policy of limiting the scope of arbitral claims to support its narrow reading, however, the tribunal rejected this argument and stressed that its task was to interpret the treaty in accordance with the VCLT, not to align it with one party’s alleged policy (para. 261).

Can the MFN Clause Be Relied Upon to Expand Arbitral Jurisdiction?

In arbitral practice, investors have frequently attempted to expand the scope of jurisdiction by invoking the MFN clause, but have usually done so without success.¹⁵ The Junefield case was no exception. The tribunal began by framing the issue: whether the MFN clause could expand jurisdiction beyond Article 9.3 to include FET and PS claims (para. 277). It first examined the treaty text and noted that Article 3.2—where the MFN clause is located—refers only to “treatment and protection” as defined in Article 3.1, which, in turn, refers to FET and protection, with no mention of dispute resolution. Applying the VCLT, the tribunal found that these terms—while undefined—do not, in their ordinary meaning or context, imply a right to arbitration.¹⁶ It rejected Junefield’s authorities and arguments, emphasizing that arbitration is a procedural mechanism, not a substantive standard of treatment (para. 289).¹⁷ Looking to Article 9.3, the tribunal underscored that arbitration was deliberately limited to compensation for expropriation. Expanding it through MFN would contradict the parties’ express agreement (para. 295). Citing Daimler v. Argentina ¹⁸ and the World Bank Guidelines, the tribunal affirmed that “treatment” was never understood to include access to arbitration. It thus upheld Ecuador’s objection, finding that the MFN clause could not be used to expand jurisdiction to FET and PS claims.¹⁹

In conclusion, the tribunal held, by majority, that Article 9.3 conferred jurisdiction not only to determine the amount of compensation for expropriation but also to assess whether an expropriation occurred and whether it resulted in a treaty breach. However, the tribunal also unanimously held that jurisdiction—whether under Article 9.3 or through the invocation of the MFN clause—did not extend to claims involving other substantive protections under the BIT, such as FET and PS.

* * * *

The Junefield award reflects the tribunal’s deliberate effort to challenge the path of narrowly interpreting restrictive ISDS clauses in early-generation BITs. It serves as a cautionary signal for China and other countries with similarly narrow worded BITs: restrictive ISDS clauses do not necessarily preclude jurisdiction altogether,²⁰ while ambiguous treaty language may give rise to broader obligations than originally intended.

The award offers clear strengths and valuable insights into the interpretation of restrictive ISDS mechanisms. It underscores the primacy of treaty text and a neutral interpretive standard, which reflects the tribunal’s faithful application of the VCLT. In practice, the precise wording of restrictive ISDS clauses and the context of the treaty are critical. In Junefield, the tribunal closely examined the specific BIT language—particularly the phrase “involving the amount of compensation for expropriation”—through both textual and contextual analysis (paras. 225–48). It also engaged in comparative treaty interpretation (paras. 250–56). Although restrictive clauses were common in China’s BITs from the 1980s and 1990s—reflecting its policy at the time—the tribunal adhered to textual analysis and avoided favoring either party’s historical policy stance. The award thus offers two key warnings. First, the specific phrase and seemingly minor terms like involving, relating to, or arising out of can carry significant legal consequences. Second, treaty drafters should ensure consistency and coherence across BITs to minimize conflicting interpretations in future disputes.

While the tribunal faithfully applied textual and contextual interpretation, it arguably overlooked a critical issue—the treaty’s drafting history and the original intent of the parties. These are expressly recognized under the VCLT as relevant interpretive tools.²¹ Notably, this omission was also one of the central concerns raised by the dissenting arbitrator.²² Historically, many socialist countries, including the former Soviet Union, Eastern European countries, and China, adopted restrictive ISDS clauses in their BITs to assert sovereignty and resist international arbitration.²³ The roots of China’s approach trace back to its 1990 accession to the ICSID Convention, when China limited consent to arbitration strictly to disputes “involving the amount of compensation for expropriation,” leaving other matters to domestic courts.²⁴ While expressed differently across treaties, this reservation defined China’s first- and second-generation BITs.²⁵ Yet the Junefield tribunal majority prioritized “the most harmonious interpretation” (emphasis added, para. 261) and maintained that its task was not to align the treaty with a party’s historical policy preferences. Giving decisive weight to China’s historical intent would have required abandoning the broader interpretation adopted—something the tribunal majority may have consciously avoided. In this light, the decision appears less surprising than inevitable.

The award also highlights a deeper structural dilemma: restrictive ISDS clauses place China in a position of conflicting interests. As a respondent, China benefits from narrow interpretations that shield the state. But broad readings better protect Chinese investors. The real challenge, however, lies in inconsistency—both in the language of China’s BITs and in arbitral jurisprudence. Treaty texts vary, and tribunals, though text-focused, often interpret similar clauses differently. Therefore, divergent outcomes in comparable cases are increasingly evident.²⁶ As China’s outbound investment grows rapidly, its investors are turning more frequently to ISDS claims. But restrictive clauses still dominate Chinese BITs. Investors often rely on these clauses to establish jurisdiction—only to see tribunals adopt narrow readings, as in Beijing Shougang v. Mongolia and Beijing Everyway Traffic v. Ghana, where claims were dismissed. This creates persistent jurisdictional uncertainty for Chinese investors.

Looking ahead, that uncertainty is already unfolding. Currently, four China-related ISDS cases—three initiated by Chinese investor and one against the Chinese government—are pending before ICSID and the PCA,²⁷ all involving attempts to establish jurisdiction based on restrictive ISDS clauses, and more are likely to follow. China’s position remains difficult: it advocates for narrower interpretations to safeguard its government, yet broader ones to benefit its investors. Nonetheless, arbitral decisions are not always predictable or favorable. The Junefield award, however, offers a positive signal for both Chinese and foreign investors, as it provides future tribunals a reference point for adopting a broader reading of restrictive clauses. Notably, PowerChina and others v. Viet Nam (II)—one of the four pending China-related cases referenced earlier—is based on the 1992 China-Vietnam BIT, whose restrictive ISDS clause and the contextual structure are nearly identical to those of the 1994 China-Ecuador BIT in Junefield. Whether the PowerChina tribunal will follow Junefield’s broad interpretation is thus a matter of significant interest. More broadly, cases like Junefield highlight the pressing need for China and other treaty parties to revise and refine their treaty drafting practice. They provide a compelling justification for adopting more precise, coherent language in the next generation of BIT negotiations.