1.1 A brief history of the modern SDR

Sovereign debt¹ restructurings (“SDR”) have undergone significant changes in the long history of the sovereign debt crisis.² The modern SDR is rooted in the Bretton Woods Conference, which initiated the International Monetary Fund (“IMF”),³ although current SDRs differ dramatically from those at that time. The Paris Club provided the framework for restructuring external bilateral sovereign debt, when sovereign debt of a state was owned by other governments.⁴ The rapid growth of private lending to developing countries in the late 1970s resulted in the emergence of a negotiating procedure for the restructuring of commercial bank debt called the London Club process.⁵ Although the London Club process required unanimity to conclude a deal, i.e. changes to negotiated payment terms, it is observed that in the 1970s “[c]ases of holdout litigation against the debtors were very rare”.⁶ This changed in the 1980s, however. For example, in Allied Bank International v. Banco Credito Agricola de Cartago, the U.S. Court of Appeals held that as a matter of interpretation of United States policy, in SDR procedure, “while parties may agree to renegotiate conditions of payment, the underlying obligations to pay nevertheless remain valid and enforceable”.⁷ The court thus confirmed the possibility that a minority of creditors might “hold out”.⁸ Moreover, a major policy shift into the nature of sovereign debt was introduced in 1989, by the so-called Brady Plan.⁹ The essence of this plan includes the exchange of outstanding bank loans into sovereign bonds as tradable instruments. The most significant impact of this policy shift is that it brought changes to the creditor structure. As bonds can be sold to retail investors, they may be held by a dispersed group of creditors.¹⁰ For the restructuring of sovereign bonds, a bond exchange is used, the success of which depends on the participation rate by bondholders.

1.2 Widespread adoption of CACs

In the course of these changes to the structure and nature of SDR, various proposals have been made to achieve swift and orderly SDR.¹¹ Of these, the two major proposals are the “statutory proposals”, aimed at the creation of sovereign bankruptcy mechanisms; and the “contractual approach”, which attempts to address sovereign debt problems by improving the terms of individual bonds.¹² The former has not seen the light of day – the “Sovereign Debt Restructuring Mechanism” (“SDRM”) proposed by IMF management and staff in 2002 failed to obtain political support and has not yet been launched. In its place, the contractual approach, which does not require statutory reform, has been widely adopted.¹³ One of the innovations of the contractual approach is the inclusion of collective action clauses (“CACs”) in bond contracts.¹⁴ While the inclusion of CACs addresses narrower scope of problems than the statutory approach,¹⁵ it does address the issue shared by all the proposals to improve the SDR process – the “concern with a creditor collective action problem in some form”.¹⁶ That is, CACs allow a specified supermajority of creditors to impose changes to the restructuring terms on a dissenting minority, or holdout creditors. Sovereign bonds issued under English law, New York law¹⁷ and Japanese law¹⁸ generally include CACs.¹⁹ By contrast, most sovereign bonds of EU countries (other than international bonds under English law) had not included CACs until recently. However, in November 2010, the Eurogroup announced a proposal to require the inclusion of CACs for all euro-zone sovereign debt securities issued after June 2013 (later accelerated to January 2013).²⁰ On 18 November 2011, the Economic and Financial Committee (“EFC”) agreed upon standardized and identical CACs (“standardized CACs”).²¹ According to the standardized CACs, “the terms and conditions of the Bonds or of any agreement governing the issuance or administration of the Bonds” may be modified by a supermajority (for reserved matters) or a simple majority (for non-reserved matters), and “[a]ny conversion or exchange undertaken to implement a duly approved modification will be binding on all Bondholders”.²² It follows that, where the terms are modified by the exercise of CACs, holdout creditors may not invoke the original bond terms for their portion of the claims. CACs are thus expected to reduce the litigation risk from, and other disruptive influence of, holdout creditors.²³ In the current situation, where statutory reform is not likely to be forthcoming, the contractual approach, especially with CACs, is widely accepted as the best feasible option at this juncture, as indicated by the nearly universal adoption of CACs by the major economies.

1.3 ‘Holdout’ investment treaty arbitration

Meanwhile, SDR faces a new challenge in the sphere of the law on foreign investment that bears on the claims of holdout creditors. Following Argentina’s default on its foreign debt in late 2001, a number of Italian holders of the defaulted bonds filed claims against Argentina under the Argentina/Italy bilateral investment treaty (“BIT”). Three investment arbitration tribunals were constituted under the International Centre for Settlement of Investment Disputes (“ICSID”) to hear such claims,²⁴ two of which issued decisions on jurisdiction and admissibility in 2011 (Abaclat v. Argentina) and 2013 (Ambiente v. Argentina). In both of these cases, the sovereign bonds in question had been issued in international markets under laws other than that of Argentina such as New York law, and the claimants bought these bonds through secondary market transactions. The common characteristic of these cases is that they were multi-party proceedings in which the interests of the claimants were represented by a representative (l’Associazione per la Tutela degli Investitori in Titoli Argentini (“TFA”)²⁵ for the Abaclat case, and North Atlantic Société d’Administration (“NASAM”)²⁶ for the Ambiente case). While the claimants’ bonds were governed by different laws and subject to jurisdiction of the court of different countries, they all had common features: the claimants were (allegedly) Italian nationals; and there was a BIT between Argentina and Italy. Thus, multi-party proceedings were possible (insofar as the tribunal found it to be permissible under the ICSID Convention), if the Claimants invoked claims under the BIT.²⁷ This might explain why they had recourse to investment treaty arbitration – investment treaty arbitration was a “handy” forum for the TFA and NASAM, which represent holdout creditors, because it could encompass all of their claims regardless of the difference in bond terms.

For the first time in the history of investment arbitration, the gate was opened for holdout creditors of international sovereign bonds in both tribunals, by their concluding that the general prerequisites for jurisdiction had been met and that the claims were admissible. However, both decisions are accompanied by strong dissenting opinions by Professor Abi-Saab (Abaclat) and Judge Bernárdez (Ambiente), an indication of the controversial nature of these decisions.

As will be examined below (Section 4), the use of investment treaty arbitration in the Abaclat and Ambiente cases is neither supported by the Argentina-Italy BIT nor consistent with the object and purpose of investment treaties. More importantly, the approach adopted by these tribunals has the effect of undermining the primary purpose of CACs, i.e. to facilitate the SDR process by reducing the risk of holdout litigation (see Section 3.1). This article addresses the potentially disruptive effect of the expansive use of investment arbitration in the SDR process. The underlying theory is that a swift and orderly SDR also serves the purpose of investment treaties to enhance the economic development of states and therefore the investment treaty regime should respect the need to conduct the SDR process in an orderly and efficient manner. In other words, while investment treaties provide protection of foreign investment that goes beyond contractual protection, such protection must be balanced against the need for sovereign debt renegotiation.

1.4 Structure of this article

Against this background, this article examines the limits on the use of investment treaty arbitration in the context of SDR. A summary of the Abaclat and Ambiente decisions is presented first (Section 2), followed by a section demonstrating that, while the bonds at issue in these cases did not include CACs, the reasoning adopted by the majority of these tribunals has the effect of neutralizing the function of CACs (Section 3). Having thus demonstrated that these tribunals’ approach may lead to a protracted sovereign debt crisis, the discussion then examines the negative impact of the delay in the SDR process on both the creditors and the debtor state. Having demonstrated the problems caused by an excessive/improper use of investment arbitration in this context, Section 4 proceeds to examine how the limits on such use should be determined, focusing on the nature of the debtor state’s acts in the SDR process. The essence of the argument is as follows. While liabilities under most investment protection obligations arise from the exercise of sovereign power of the host state, defaults and renegotiations of international bonds and/or the implementation of CACs do not constitute acts of a sovereign, but rather, commercial activities. Such acts therefore are incapable of constituting a violation of the treaty obligations. Accordingly, in light of the prima facie standard for jurisdiction, disputes arising out of such acts lack jurisdiction ratione materiae, i.e. subject matter jurisdiction, in most cases. This article concludes by arguing that the clear recognition of/respect for the limits of investment treaty arbitration in this context helps both SDR and the investment treaty regime to operate in harmony, without compromising each other’s interests.

2.1 Abaclat and others v. Argentina

This case stems from Argentina’s default over US $100 billion of international bond in December 2001. From 1991 to 2001, Argentina issued 179 sovereign bonds in international capital markets, 173 of which were denominated in foreign currency. The claimants are (allegedly) holders of 83 of the 173 foreign currency bonds.²⁸ In 2005, Argentina launched the first exchange offer (“Exchange Offer 2005”).²⁹ In the same year, it enacted Law 26,017 (“Emergency Law”) which provides, inter alia:

…with regard to those bonds which were eligible for but were not exchanged in the Exchange Offer 2005 (i) the Executive Branch of the government shall not reopen the exchange process; and (ii) the national government is prohibited from entering into any juridical, extra-juridical or private transaction.³⁰

In the United States, a number of court actions were initiated by creditors who were unsatisfied with the terms and conditions of Exchange Offer 2005. According to Argentina, in two of these actions, many of the plaintiffs were the claimants to the Abaclat arbitration.³¹ In 2006, a U.S. law firm filed a Request for Arbitration with ICSID on behalf of over 180,000 Italian bondholders.³² Subsequently, a considerable number of the claimants tendered into the second exchange offer made in 2010 (“Exchange Offer 2010”) and withdrew from arbitration. As a result, on the date of the Decision on Jurisdiction and Admissibility, the number of remaining claimants was approximately 60,000.

Argentina contested the tribunal’s jurisdiction on various grounds. For the purposes of this article, the most pertinent is that the claims are not treaty claims, but rather, contract claims “for which the relevant contractual instruments provide non-Argentine legal rights and remedies that could not be and were not affected by any act of Respondent”.³³

On 4 August 2011, the tribunal issued its decision on jurisdiction and admissibility, in which the majority concluded that, while leaving the examination of jurisdictional issues in relation to individual claimants to the merits phase, the general prerequisites for the jurisdiction of ICSID tribunals had been met, and the claims were admissible.

The majority rejected Argentina’s objection that the claims may not be considered as treaty claims, stating that the facts alleged by the claimants, if established, were “susceptible of constituting a possible violation of at least some of the provisions of the BIT invoked by the claimants”.³⁴ Likewise, the majority rejected Argentina’s argument that the claims were purely contractual, stating that the promulgation of the Emergency Law entitling it not to perform part of the obligations under the bonds:

…derives from Argentina’s exercise of sovereign power. Thus, what Argentina did, it did based on its sovereign power; it is neither based on nor does it derive from any contractual argument or mechanism.³⁵

The majority thus concluded that, as a prima facie case, the claims presented in the case were treaty claims based on acts of a sovereign.³⁶

2.2 Ambiente and others v. Argentina

In February 2013, another ICSID tribunal issued the decision on jurisdiction and admissibility on the case stemming from Argentina’s sovereign debt default in 2001. The facts³⁷ and arguments presented by Argentina before the Ambiente tribunal³⁸ were very similar to those of the Abaclat case, except that the number of the claimants was substantially smaller in Ambiente than in Abaclat.³⁹

In its decision on jurisdiction and admissibility, the Ambiente tribunal reached the same conclusion as the Abaclat tribunal. With respect to the issue of whether or not the claims constituted treaty claims, the tribunal provided a more detailed analysis than its “sister tribunal”⁴⁰ had in the earlier case. In Ambiente, Argentina argued that it acted not in the exercise of its sovereign authority, but merely as a commercial party, on the grounds, inter alia, that: (a) it was impossible for it to exercise sovereign authority in regard to the claimants’ security entitlements that were governed by foreign law and beyond the scope of Argentina’s legislative jurisdiction and (b) non-payment in itself was no violation of international law and therefore may not constitute a basis for treaty claims.⁴¹

The tribunal first rejected the latter argument, stating that:

…it was not so much the failure to pay, but the use of the Respondent’s sovereign prerogatives when restructuring its debt … which qualify the Respondent’s acts as potential breaches of the Argentina-Italy BIT and thus as treaty claims.⁴²

The tribunal also rejected the former argument:

[I]nsofar as the Respondent seeks to conclude from the existence of such choice of law and forum selection clauses that those instruments were, by definition, beyond the scope of Argentina’s legislative jurisdiction, the present Tribunal cannot follow this reasoning. While the Respondent could obviously not alter the terms of legal rights and obligations as arising from different laws and jurisdictions, it could nonetheless influence those bonds/security entitlements within the reach of the Respondent’s (notably territorial) jurisdiction, for instance by legally forbidding the executive authorities to enter into any settlement of the claims in question or by ordering the domestic judicial authorities, should an “old” bond come before them, to replace ipso jure the old bonds by the newly issued bond instruments.⁴³

The tribunal, applying the prima facie test for jurisdictional purposes,⁴⁴ therefore concluded that Argentina’s acts “can plausibly be understood as having unilaterally altered the contractual equilibrium and having transcended the realm of purely non-sovereign action”, which may be capable, if proved, of constituting breaches of the BIT.⁴⁵

3.1 The potential effect of the Abaclat and Ambiente approaches on CACs: treaty claims vs. contract claims

In investment treaty arbitration, it is well established that treaty claims may arise from a contractual relationship between foreign investors and the host state, independently of the existence of a breach of the contract. The tribunal in Vivendi v. Argentina II stated that:

Articles 3 (fair and equitable treatment) and 5 (expropriation) of the BIT do not relate to breach of a municipal contract. Rather, they set an independent standard. A state may breach a treaty without breaching a contract; it may also breach a treaty at the same time it breaches a contract.⁴⁶

This distinction between a contract claim and a treaty claim has been accepted by a number of tribunals.⁴⁷ Treaty claims do not arise if the investor’s claim is based solely on the alleged breach of the contract by the host state. The Abaclat tribunal itself stated that “[a] claim is to be considered a pure contract claim where the Host State, party to a specific contract, breaches obligations arising by the sole virtue of such contract”.⁴⁸

The only exception to this is the so-called umbrella clause, which requires a host state to respect any obligation assumed by it with regard to a specific investment. It is largely accepted that a broadly drafted umbrella clause “elevates” a breach of contracts, including commercial contracts, to the level of a treaty breach.⁴⁹ Applying this to the SDR process, however, in situations where the relevant bond includes CACs and a supermajority agrees to accept the restructuring plan, this “treaty claim” will also be barred. This is because, in such cases the modification of the terms and conditions of the bonds is legally binding on all creditors by the application of the CAC, and accordingly, holdout creditors may not raise contractual causes of action for full payment. In other words, insofar as the obligations under the contract are concerned, and the debtor state implements the CAC in good faith, it is highly unlikely that the debtor state’s “failure to pay” constitutes a violation of the umbrella clause. The United Nations Conference on Trade and Development (“UNCTAD”), in its report on the relationship between SDR and investment treaty arbitration, aptly observed that “it would appear that … dissenting bondholders cannot succeed in their IIA (International Investment Agreement) claims, given that their contractual rights have been duly modified”.⁵⁰

Therefore, the identification of the host state’s acts relating to sovereign debt default and restructuring that form the basis of the investor’s claim is crucial. If the tribunal finds that the holdout creditors’ claims are based solely on the debtor state’s default and renegotiation of bond terms (1) there will be no basis of treaty claims apart from an umbrella clause claim and (2) the umbrella clause claim, even if invoked, will be barred by virtue of any successful implementation of CACs. Yet this is not what the Abaclat and Ambiente tribunals did. Both tribunals considered that Argentina’s acts that were alleged by the claimants to be violations of treaty claims were beyond Argentina’s default and renegotiation of bond terms.⁵¹ The Abaclat tribunal found that not only Argentina’s failure to perform its payment obligations but also its intervention as a sovereign to “justify its failure” and to “modify its payment obligations towards its creditors”, constituted potential breaches of the BIT.⁵² In particular, it stated that “[t]he arbitrary promulgation and implementation of regulations and laws can … under certain circumstances, amount to an unfair and inequitable treatment”.⁵³ The Ambiente tribunal went further, stating that “it was not so much the failure to pay, but the use of the Respondent’s sovereign prerogatives when restructuring its debt … which qualify the Respondent’s acts as potential breaches of the Argentina-Italy BIT and thus as treaty claims”.⁵⁴ Both tribunals therefore identified the enactment and implementation of regulations and laws such as Law 26,017 as acts that were independent of Argentina’s conduct as a party to the bond contracts, thereby constituting potential breaches of treaty obligations. According to these tribunals’ identification of the acts that gave rise to the claims, irrespective of the contractual position of the debtor state under the relevant bond, the state may still be found responsible for a breach of treaty obligations.⁵⁵ It follows that, even where the bond includes CACs and a supermajority agrees to accept the restructuring plan, holdout creditors, who are bound by such decision as a matter of contract law, may still be able to resort to investment treaty arbitration by claiming that the debtor state exercised its sovereign power in the SDR process and it amounts to a breach of treaty obligations. This approach therefore does leave treaty claims “untouched” by CACs.⁵⁶

Against this backdrop, the remainder of this section first examines how such an approach disrupts the SDR process. It then demonstrates the negative consequences of the delay in the process for the creditors and the debtor countries.

3.2 Investment arbitration and the creditor coordination problem

As noted earlier, CACs are designed to address the creditor coordination problem of reaching agreement among creditors. In the process of sovereign bond restructuring, the borrowing country makes a “take-it-or-leave-it” offer to exchange the existing bonds for new ones with less favourable terms than the original bond terms.⁵⁷ Holdout risks arise in this phase – in particular, the risk manifests itself in the form of litigation initiated by holdout creditors “who oppose a settlement hoping to secure better terms for their portion of the claims at the expense of the rest of the creditor group (free riding)”.⁵⁸ It is observed that the “threat of being tied up for years in litigation … might encourage sovereigns to pay holdouts, which will in turn encourage parties to hold out and make restructurings harder to achieve”.⁵⁹ The creditor coordination problem thus causes significant delay in the SDR process. To address this problem, CACs bar contractual causes of action by holdout creditors, thereby effectively reducing the risk of holdout litigations.

To be sure, it remains controversial whether or not CACs actually expedite the SDR process.⁶⁰ Still, the benefit of avoiding the holdout problem for debt renegotiations is clear. At the least, it eliminates an important factor that discourages intra-creditor strategic play that results in unfair/unequal treatment.⁶¹ The almost universal adoption of CACs for international sovereign bonds (see Introduction) suggests that CACs are considered to be an effective means to achieve a swift and orderly SDR⁶² under current conditions, where a comprehensive regime for SDR does not exist. However, if investment arbitration by holdout creditors has potential to surpass the contractual restrictions set by CACs (as examined above), the creditor coordination problem continues to exist. Moreover, while the terms of the offer have to be more attractive than “the alternatives faced by creditors”⁶³ in order for the “take-it-or-leave-it” exchange offer to invite wide participation, investment arbitration offers a very attractive “alternative” to creditors. This is because investment arbitration awards, in particular ICSID arbitration awards, are more advantageous than municipal judgments in terms of enforcement. Enforcing a local court’s judgment against a state presents obstacles (as explained below). By contrast, there is a strong enforcement mechanism in the ICSID arbitration regime: Article 54 of the ICSID Convention provides that each contracting state must enforce the pecuniary obligations imposed by ICSID arbitration awards “within its territories as if it were a final judgment of a court in that State”.⁶⁴ Certainly, state immunity from execution of assets is preserved even under the ICSID Convention: Article 55 of the Convention provides that “[n]othing in Article 54 shall be construed as derogating from the law in force in any Contracting State relating to immunity of that State or of any foreign State from execution”. Yet in practice, the rate of compliance with ICSID awards is very high⁶⁵ – indeed, in the ICSID arbitration system “only a small number of awards have not been complied with”⁶⁶ (important exceptions are Russia and Argentina).⁶⁷ Thus, if holdout creditors may still resort to investment arbitration and obtain full and practically enforceable compensation, this will undoubtedly discourage creditors from participating in a restructuring offer. Creditors will not be willing to take a loss knowing that holdouts may receive full payments elsewhere.

3.3 Consequences for the creditors and the debtor states

There is little doubt that defaults and debt restructuring are costly for both the creditors and the borrower,⁶⁸ and delays in the SDR process causes negative consequences. Benjamin and Wright examined haircuts for all countries that defaulted in the period 1980–2004 and concluded that longer delay is associated with “larger haircuts”.⁶⁹ Reinhart and Rogoff observe that “international debt reschedulings typically saddle investors with illiquid assets that may not pay off for decades”.⁷⁰ Tsatsaronis explains the costs incurred by creditors as a result of prolonged negotiations as follows:

Investors typically see the secondary market value of their claims plummet during the renegotiation period, with distressed securities specialists representing the sole source of liquidity in the market. The longer a bond spends in default, the lower its recovery rate.⁷¹

It is also observed that the harms caused to the creditors by a default are “made up by positive returns in normal times”.⁷² This indicates that returning debt to sustainable levels – allowing the debtor state to experience normal times – by accepting a workable debt deal, ultimately serves the interests of the creditors.⁷³ To be sure, the creditors’ conduct is shaped by different strategies. For example, Pitchford and Wright observe that, in an unstructured negotiation environment, a strategic motivation to “wait until others have settled, in an attempt to extract a larger settlement” arises. It should be noted, however, that the inclusion of CACs is “one policy response” to address the problem of an unstructured negotiation environment.⁷⁴

As to default costs to the debtor, Sturzenegger and Zettelmeyer identify categories of costs that have empirical support: loss of capital market access, adverse effects on international trade, borrowing costs and reputational spillovers.⁷⁵ First, when a country defaults on its debts, it is practically excluded from the market until it reaches a satisfactory arrangement in negotiation with the creditors of such debts.⁷⁶ Secondly, a country in default will be “forced to conduct its trade in roundabout ways to avoid seizure”.⁷⁷ The same will apply to foreign direct investments, which are by nature transnational. Lenders may cut off the country’s access to trade credit during the period of default.⁷⁸ Thirdly, it is very difficult and expensive for the country to obtain new loans if there remains the possibility that its credit capacity is materially affected by claims by holdout creditors.⁷⁹ Lastly, the loss of confidence in the debtor state caused by a default will lead to capital flight and a large reversal of inflows.⁸⁰ Protracted SDR process thus prevents the country in default from resolving its financial crisis and restore financial, economic and political stability that were undermined by sovereign debt crisis.

4.1 The commercial activities of the debtor state

This section first demonstrates that Argentina’s international sovereign bonds at issue in the Abaclat and Ambiente cases should be excluded from the scope of investment treaty protection at the jurisdictional phase for the following reasons. While most investment treaty obligations may be breached only by sovereign acts of the host state, default and restructuring of these international bonds do not constitute acts of a sovereign, and therefore are “incapable of falling within” most investment protection obligations. Accordingly, in light of the prima facie standard for jurisdiction, the claims lack jurisdiction ratione materiae. It then argues that even where domestic bonds are at issue, insofar as the debtor state implements the CACs contained in the relevant bonds in SDR, such acts of the state do not constitute the acts of a sovereign in relation to the bonds, and the tribunal should therefore reach the same conclusion with regard to international sovereign bonds.

To demonstrate these arguments, the following section first demonstrates the “sovereign acts” requirement for a breach of investment protection obligations (Section 4.1.1). Sections 4.1.2 and 4.1.3 examine the distinction between sovereign acts and commercial activities. Section 4.1.2 considers how criteria for the distinction have been developed in the law of state immunity. Section 4.1.3 then turns to the question of whether/how such criteria should also be applied in investment treaty arbitration. It will demonstrate that the criteria widely accepted in the law of state immunity have also been adopted in investment arbitration in the context of distinguishing between contract claims and treaty claims. On the other hand, the “borrowing” of the criteria comes with the proviso that the purpose and context of distinguishing these two types of acts in investment treaties is different from those in the law of state immunity, and therefore, when relying on these criteria, factors particular to investment treaties must be taken into account. Having thus established the criteria applicable to investment treaties, Section 4.1.4 applies the criteria to the debtor state’s sovereign debt default and restructuring. It first describes how such criteria have been applied by domestic courts to the cases of sovereign debt default. For this purpose, the current analysis focuses on English and U.S. courts as the two jurisdictions most often chosen for international sovereign bonds.⁸⁴Section 4.1.5 proceeds to demonstrate that the application of the (modified) criteria to investment treaties leads to the conclusion that Argentina’s acts at issue in the Abaclat and Ambiente cases with regard to international bonds do not amount to the acts of a sovereign, but remain in the realm of commercial transactions. It also argues that in cases of SDR of domestic sovereign bonds, insofar as the debtor state implements the CACs contained in the relevant bonds, such acts also remain in the realm of commercial transactions.

4.2 The prima facie test for jurisdiction

It is well established that for jurisdiction ratione materiae (subject matter jurisdiction) to be established, the tribunal must satisfy itself that the facts alleged by the claimant, if ultimately proved, would be capable of “falling within (or coming within) (or constituting a violation of) the provisions of the investment treaty” (the prima facie test).¹³⁴ In other words, the tribunal must satisfy itself at the jurisdictional stage that the connexion between the alleged facts and their legal characterization, i.e. the legal foundation of the claims, is established.¹³⁵ The tribunal in Pan American Energy v. Argentina provides a strong reason for this proposition:

…if everything were to depend on characterisations made by a claimant alone, the inquiry to jurisdiction and competence would be reduced to naught, and tribunals would be bereft of the compétence de la compétence enjoyed by them under Article 41(1) of the ICSID Convention.¹³⁶

Applying this objective examination on the legal foundation to the Abaclat and Ambiente cases, the appropriate conclusion would be that, insofar as the investment protection obligations (save the umbrella clause) are concerned, Argentina’s acts as commercial activities are incapable of falling within these provisions and therefore the disputes lack subject–matter jurisdiction.

This is not to argue that any disputes arising from SDR are always excluded from the scope of investment arbitration. The subject matter jurisdiction may be established with regard to disputes over the host state’s acts relating to SDR in exceptional circumstances where, for example, the relevant BIT explicitly includes sovereign debts (see introduction of Section 4) and the dispute involves an exercise of the debtor state’s sovereign power.¹³⁷ Yet it should be emphasised that the establishment of subject–matter jurisdiction does not mean that the state’s acts are found to constitute a breach of investment treaty obligations at the merits phase.¹³⁸ This naturally requires a fact-specific assessment, a detailed examination of which is a subject for future study. It is, however, worth emphasising that in relation to CACs, investment arbitration tribunals, in conducting this assessment, should appreciate that excessive interference by investment arbitration with a carefully structured SDR process may well undermine the future implementation of CACs. When the SDR process is conducted in good faith, investment treaty arbitration should not be used as a side door by holdout creditors seeking to evade the contractual restrictions posed by the CAC.